Hydrogénation of Aliphatic Nitriles over Nickel Catalysts Modified by ...

Research Collection

Doctoral Thesis

Hydrogenation of aliphatic nitriles over nickel catalysts modifiedby formaldehyde

Author(s): Novi, Roc

Publication Date: 2004

Permanent Link: https://doi.org/10.3929/ethz-a-004877494

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.

ETH Library

https://doi.org/10.3929/ethz-a-004877494

http://rightsstatements.org/page/InC-NC/1.0/

https://www.research-collection.ethz.ch

https://www.research-collection.ethz.ch/terms-of-use

DISS. ETHNO. 15708

Hydrogénation of Aliphatic Nitriles over

Nickel Catalysts Modified by Formaldehyde

A dissertation submitted to the

SWISS FEDERAL INSTITUTE OF TECHNOLOGY ZURICH

for the degree of

Doctor of natural sciences

presented by

ROC NOVI

Dipl. Chem. ETH

born on 15. October 1977

citizen of Vignogn GR

accepted on the recommendation of

Prof. Dr. P. Rys, examiner

Prof. Dr. M. Morbidelli, co-examiner

Zürich 2004

Page h

Acknowledgements

I wish to express my sincere gratitude to...

...Prof. Dr. P. Rys, who gave me the chance to work on this project. I am

very grateful to him for many discussions and his continuous support.

...Prof. Dr. M. Morbidelli for accepting co-examination of this doctoral

thesis.

...Dr. F. Rössler for the possibility to perform the experiments in

Kaiseraugst, the scientific support and the discussions. His confidence in

my work was very supporting and encouraging.

...Dr. A. Rössler and B. Sägesser for the numerous discussions, tips and

advices.

...my colleagues of the Rys group, especially Dr. F. Antognoli,

Dr. E. Dedeoglu, Dr. A. J. Klaus, Dr. M. Mösche, Dr. A. Rössler,

D. Schoch and Dr. P. Skrabal.

...the members of the hydrogénation team in Kaiseraugst, H. Bruder,

B. Close, A. Dodane, H. Lehmann, T. Müller, A. Saaler, B. Sägesser,

R. Santillo, P. Schmidt, G. Weisser and C. Zürcher for the good team

work.

...the members of the analytical groups in Kaiseraugst and Zürich

H. Kleissner, Dr. G Schiefer and Dr. P. Skrabal for the coaching and

measurements they performed.

...M. Bäbler, L. Brändli, M. Bucher, S. Diezi, M. Günther, J. Fischesser,

P. Müller and S. Sasso for the discussions and practical advice.

Page m

.my friends, D. Günther, I. Netzer, M. A. Plaz and J. W. Solèr for their

friendship and moral support apart from the work.

.my parents, my grandparents as well as my brother and sister for their

support during all these years of education.

.the company F. Hoffmann-La Roche for the financial support of this

project as well as the possibility to perform the experiments in its

laboratory in Kaiseraugst.

Page lv

Table of contents

1 Abstract 1

2 Zusammenfassung 3

3 Introduction 5

3.1 Overview 5

3.2 Aim and scope of this thesis 7

3.3 List of abbreviations and symbols 8

3.4 List of substances 9

4 Theoretical section 13

4.1 Hydrogénation of nitriles 13

4.1.1 General aspects 13

4.1.2 Hydrogénation of nitriles to amines 13

4.1.3 Hydrogénation of nitriles to aldehydes 15

4.1.4 Hydrogénation of nitriles to hydrocarbons 16

4.1.5 Hydrogénation and cyclisation 16

4.2 Mechanistic considerations of nitrile hydrogénation 17

4.2.1 Historic development 17

4.2.2 Models for the formation of side products 23

4.2.3 Influence of reaction parameters on nitrile hydrogénation 25

4.3 Reversible reactions of amines 27

4.4 Raney nickel 30

4.4.1 Preparation methods 30

4.4.2 Industrial applications 31

4.4.3 Variation of the properties of Raney nickel 32

4.4.4 Adsorption of nitriles and their hydrogénation

intermediates 33

4.4.5 Inhibition/poisoning of the catalyst 35

4.4.6 Acid sites 36

5 Hydrogénation of butyronitrile 39

5.1 Reaction system 39

Page v

5.2 Thermodynamic aspects 45

5.3 Aspects of mass and heat transport 45

5.4 Influence of reaction parameters on nitrile hydrogénation 47

5.4.1 Influence of the overall pressure 48

5.4.2 Influence of the temperature 50

5.4.3 Influence of the ratio of catalyst to substrate 54

5.4.4 Recycling of the catalyst 57

5.4.5 Influence of additives 59

5.5 Influence of washing/modification procedures 60

5.5.1 Influence of washing procedures with different solvents 60

5.5.2 Influence of the modification by formaldehyde 63

5.6 Reversibility of the hydrogénation steps 68

5.6.1 Disproportionation of butylamine 68

5.6.2 Influence of reaction parameters on the disproportionation

of butylamine 70

5.6.3 Dibutylimine as starting material 72

5.7 Discussion 73

5.7.1 The bifunctional catalytic hydrogénation and its

reversibility 73

5.7.2 Influence of various reaction parameters on the selectivity ...74

6 Modification of nickel catalysts by formaldehyde 77

6.1 General remarks 77

6.2 Influence of the treatment with different solvents on the

properties of the catalyst 79

6.2.1 Reduction potential 79

6.2.2 Adsorption of an indicator 80

6.3 Modification of Raney nickel by various formaldehyde

concentrations 81

6.3.1 Analysis of the modifying solution 81

6.3.2 Properties of the modified catalysts 84

6.4 Modification of various amounts of Raney nickel at constant

modification strength 85

6.4.1 Analysis of the modifying solution 85

6.4.2 Properties of the modified catalysts 87

Page vi

6.5 Modification of nickel-on-carrier 89

6.6 Discussion 89

7 Effect of formaldehyde modified nickel catalysts on other chemical systems ...91

7.1 Hydrogénation of crotonaldehyde 91

7.1.1 General remarks 91

7.1.2 Test for a possible gas-liquid transfer limitation for

hydrogen 93

7.1.3 Influence of the formaldehyde modification of Raney

nickel on the hydrogénation of crotonaldehyde 94

7.1.4 Influence of the formaldehyde modification of

nickel-on-carrier on the hydrogénation of crotonaldehyde ...96

7.2 Hydrogénation of l-bromo-4-nitrobenzene 97

7.2.1 General aspects 97


hydrogen 98

7.2.3 Influence of the modification on selectivity and reaction

rates 99

7.3 Hydrogénation of levodione 101

7.3.1 General remarks 101


hydrogen 101

7.3.3 Influence of the modification on selectivity and

hydrogénation rate 102

7.4 Discussion 103

8 Conclusions and outlook 105

9 Experimental 107

9.1 Apparatus 107

9.1.1 Description of the 500 ml steel hydrogenator 107

9.1.2 200 ml glass hydrogenator 109

9.1.3 100 ml low pressure hydrogénation apparatus 109

9.1.4 Lab Shaker 110

9.1.5 Modification and washing apparatus 110

9.1.6 Gas Chromatograph 110

9.2 Methods ill

Page vu

9.2.1 Hydrogénation of butyronitrile Ill

9.2.2 Reversibility experiments in the 35 ml screening autoclave 111

9.2.3 Hydrogénation of crotonaldehyde 112

9.2.4 Hydrogénation of l-bromo-4-nitrobenzene 112

9.2.5 Hydrogénation of levodione 113

9.2.6 Description of the sampling procedure 113

9.2.7 Neutralisation with water 114

9.2.8 Neutralisation with methanol 114

9.2.9 Neutralisation with tetrahydrofurane 114

9.2.10 Modification with formaldehyde 114

9.3 Analytics 115

9.3.1 Determination of butyronitrile, butylamine, dibutylamine

and dibutylimine with a GC method using an internal

standard 115

9.3.2 Determination of crotonaldehyde, crotylalkohol, butanal

and butanol with a GC method using an internal standard...

116

9.3.3 Determination of l-bromo-4-nitrobenzene, l-bromo-4-

aminobenzene and aniline 118

9.3.4 Determination of levodione and actinol 119

9.3.5 Methanol determination in aqueous medium with a

headspace GC method using an external standard 120

9.3.6 Formaldehyde determination in aqueous medium with an

HPLC method using an external standard 121

9.3.7 Synthesis of dibutylimine as a standard for GC

measurements 122

9.4 Identification of by-products 123

9.4.1 N-Butylbutanamide 123

9.4.2 N,N-Dibutylbutyramidine 123

9.4.3 1,1-Diethoxybutane 124

9.5 Characterisation of the catalyst 124

9.5.1 The reduction potential 124

9.5.2 Adsorption of an indicator 125

9.5.3 Dissolution in acidic medium 125

9.6 Chemicals 125

Page vin

9.7 Calculation of selectivity and reaction rates 127

9.8 Error analysis 128

9.8.1 Precision of a gas chromatographic analysis 128

9.8.2 Precision of a sample analysed with gas chromatography ...

129

9.8.3 Precision of the selectivity and the reaction rates 130

10 Literature 131

11 Appendix 137

11.1 Curriculum vitae 137

11.2Conference contributions 138

Pagex

Chapter A

Abstract

In the present work the liquid phase hydrogénation of aliphatic nitriles as

well as an industrially applicable modification procedure for hydrogénation

catalysts [1] producing higher yields of primary amines were investigated.

In a first part, by using Raney nickel as catalyst and butyronitrile as a

model substance the influence of the gas-liquid hydrogen transfer limitation

and several reaction parameters, such as the temperature, the hydrogen

pressure, the ratio of catalyst to substrate, the recyclability of the catalyst

and various additives on the hydrogénation selectivity were investigated.

The influence of these parameters is discussed with a semi-quantitative

macro-kinetic model presented within this thesis. Furthermore, the

reversibility of the reaction steps that characterise the hydrogénation system

was investigated with an intermediate product as starting material.

A new, economically interesting and easily applicable method to rise the

selectivity towards primary amines is the modification of nickel catalysts by

formaldehyde. Several parameters of this modification process were

investigated, because the desired higher selectivity is also accompanied by

an undesired loss of activity during the hydrogénation. In particular the

amount of formaldehyde used to treat the catalyst and the amount of catalyst

at a constant formaldehyde concentration were explored. The nickel

leaching during the modification as well as during the hydrogénation was

shown to be also an economically relevant factor, if a modified catalyst is

used in an industrial process.

The effect of the modified catalysts on other chemical systems was

screened by employing the following hydrogénation systems: the

hydrogénation of an oc,ß-unsaturated aldehyde, the hydrogénation of a

halogenated nitroarene and the enantioselective hydrogénation of a cyclic

Page 1

dione. The modification by formaldehyde was not beneficial in the tested

cases as the sélectivités were not enhanced. On the contrary, a decrease in

activity was observed with the modified catalysts.

Page 2

Chapter £â

Zusammenfassung

Im Rahmen der vorliegenden Arbeit wurden die Hydrierung von

aliphatischen Nitrilen in flüssiger Phase sowie ein industriell anwendbarer

Prozess zur Modifizierung von Hydrierkatalysatoren [1], welcher höhere

Ausbeuten an primären Aminen liefert, untersucht.

Anhand der Hydrierung von Butyronitril mit Raney Nickel wurde neben

dem Gas-Flüssig-Stofftransport für Wasserstoff auch der Einfluss der

Temperatur, des Wasserstoffdrucks, des Verhältnisses von Katalysator zu

Substrat, der Rezyklierbarkeit des Katalysators und von Zusatzstoffen auf

die Selektivität untersucht und mit Hilfe eines semi-quantitativen makro¬

kinetischen Modells diskutiert. Ferner wurde die Reversibilität der

Reaktionsstufen des Hydriersystems untersucht, indem ein Zwischen¬

produkt als Substrat verwendet wurde.

Die Hydrierung mit Katalysatoren, die mit Formaldehyd modifiziert

wurden, ist eine neue, ökonomisch interessante Möglichkeit zur Steigerung

der Selektivität bezüglich primären Aminen. Zur Implementierung der

modifizierten Katalysatoren sind jedoch mehrere Parameter zu untersuchen,

da die erwünschte Selektivitätssteigerung von einem unerwünschten

Aktivitätsverlust begleitet wird. Insbesondere die Formaldehyd¬

konzentration, aber auch das Katalysator-Substrat-Verhältnis sowie deren

Auswirkungen auf eine gegebene Nitrilhydrierung wurden untersucht. Es

wurde festgestellt, dass für eine industrielle Anwendung der Katalysatoren

das Herauslösen von Nickel in die Modifizierlösung wie auch in die

Hydrierlösung ein wichtiger Kosten bestimmender Faktor ist.

Die Einwirkung der modifizierten Katalysatoren auf andere chemische

Systeme wurde ebenfalls evaluiert. Hierbei wurden die modifizierten

Page 3

Katalysatoren auch bei der Hydrierung von oc,ß-ungesättigten Aldehyden,

halogenierten Nitroaromaten und Dialdehyden getestet. Es stellte sich

heraus, dass die Formaldehyd-Modifizierung für diese Systeme keinen

Vorteil in Form einer Selektivitätssteigerung bringt: Es konnte lediglich ein

Aktivitätsverlust beobachtet werden.

Page 4

Chapter +J

Introduction

3.1 Overview

The catalytic hydrogénation of nitriles leads to a mixture of primary,

secondary and tertiary amines, amides and alcohols. The economic

efficiency of this catalytic reaction to the primary amines is influenced

mainly by the by-products mentioned above. Further research is, therefore,

necessary to determine an optimal selective hydrogénation for each nitrile.

Formation of the by-products can be minimized by using a suitable catalyst,

by modification of the catalyst, or by additives such as ammonia or

alcohols. The most important catalyst for this system is Raney nickel and its

modifications, but the exact mechanism, especially the intrinsic selectivity

determining step, is presently unknown. The aim of this work is, therefore,

to obtain a deeper understanding of the mechanism and to achieve an

optimization of the process.

The selectivity obtained during nitrile hydrogénation is determined by a

strong interaction of two catalytic functions, the acid and the hydrogénation

sites. This is shown in the following reaction scheme (Figure 3-1). The ratio

of and the distance between the acid and the hydrogénation sites determine

the ratio of the reaction constants kH4/kC5, and therefore the formation of

by-products (secondary and tertiary amines). It can be assumed that the

reaction pathway via kH3, kH4 is kinetically favoured towards the reaction

pathway via kH1, kH2. Thus, a minimal number of acid catalytic sites is

favoured.

Page 5

H2/Ni H,/Nir. n R CH

© Àkm

© À+ H + H

„

V*s

V

©

r ni-

H2/Ni ©

R CH

NH - R—CH,

^H2

H,/Ni

NH,

^H3

+ RCH2NH2

h:?

^H4

R CH NH2

©

H2N CH2 R

©

R CH NH?

HN CH2 R

A

©

R CH

-NH,

©

+ H

- R—CH,

©

-NH-,

HN CH2 R

H,/Ni R CH2

H2N CH2-©

©

H

RCH2NHCH2R

tertiary amine

Fig. 3-1 : Bifunctional mechanism for the hydrogénation of nitriles.

Page 6

3.2 Aim and scope of this thesis

The aim of this work is to investigate the mechanism presented in

Chapter 3.1 and to ultimately confirm or reject its feasibility. Therefore,

investigations on the following parameters have been performed:

catalyst

catalyst modification

hydrogen pressure

additives

temperature

solvent

In addition, a recently patented modification of hydrogénation catalysts

by formaldehyde, which leads to better selectivities towards the primary

amines if aromatic nitriles are hydrogenated, is tested for aliphatic nitriles.

The completely unknown mechanism of this modification has also to be

investigated.

Page 7

3.3 List of abbreviations and symbols

Table 3-1 : List of abbreviations and symbols.

actinol (4R,6R)-4-hydroxy-2,2,6-trimethylcyclohi

BA butylamine

BN butyronitrile

BI butylimine

cat catalyst

Damide dibutylamide, JV-butylbutanamide

DBA dibutylamine

DBD dibutylamidine

DBDH+ protonated dibutylamidine

DBI dibutylimine, dibutylazomethine

EDS l,2-bis(2-hydroxyethylthio)ethane

ESTD external standard

FAMEC8 n-caprylic acid methyl ester

FID flame ionisation detector

ISTD internal standard

levodione (6R)-2,2,6-trimethylcyclohexa-l,4-dione

MeOH methanol

RaNi Raney nickel

Tamidine tributylamidine, iV,iV-dibutylburyramidine

TBA tributylamine

TFA trifluoro acetic acid

THF tetrahydrofuran

C concentration

exp experiment

Page 8

Table 3-1 : List of abbreviations and symbols.

G gas phase

kx rate constant of the reaction x

L liquid phase

P(H2) hydrogen pressure

P(NH3) ammonia pressure

r distance

ro initial reaction rate

G standard deviation

rel. G relative standard deviation

S solid phase

T temperature

3.4 List of substances

Table 3-2 : List of substances.

/ V^x/ \

-NH2 4-aminoazobenzene

CH30

4-amino-5-cyano-2-methoxy-

pyrimidine

pynitrile

Page 9


-NH, aniline

HS O

l,2-bis(2-hydroxyethylthio)ethaneSH

Br-

Br-

-NH9 l-bromo-4-aminobenzene

-NO, l-bromo-4-nitrobenzene

butanol

NH

N

butyraldehyde

butylamine

butylimine

butyronitrile

n-caprylic acid methyl ester

Page 10


crotonaldehyde

crotylalkohol

dibutylamide

NH;

N

H

NH

dibutylamidine

protonated dibutylamidine

dibutylamine

iV,iV-dibutylbutyramidine

tributylamidine

Tamidine

Page 11


(4R,6R)-4-hydroxy-2,2,6-trimethyl-

cyclohexanone

actinol

HO' R

ypyridine

tributylamine

(6R)-2,2,6-trimethylcyclohexa-l,4-

dione

levodione

Page 12

Chapter ^

Theoretical section

4.1 Hydrogénation of nitriles

4.1.1 General aspects

The reduction of organic substances with molecular hydrogen in the

presence of solid contact substances is based on heterogeneous catalysis.

Because the substances do not react with hydrogen under standard

conditions, the reaction occurs exclusively on the contact surface. Different

processes, which depend on the adsorption of reaction partners on the

catalyst, and which are not understood in detail, determine the selectivity of

such reactions. Consequently, the result of the reduction often depends on

the catalyst and the chosen conditions.

4.1.2 Hydrogénation of nitriles to amines

The hydrogénation of nitriles is one of the main methods used in industrial

chemistry for the synthesis of primary amines. The product is always a

mixture of primary, secondary and tertiary amines. The selectivity of this

catalytic reaction depends on several factors: The nature of the catalyst

(metal and support), the addition of ammonia, the temperature and the

solvent used [2-4]. This was impressively demonstrated in the case of

butyronitrile {Figure 4-1) [5].

To control the selectivity towards the primary amine, the following

methods are proposed in the literature [6] :

Page 13

Ni / diatomacous earth (50%),

NH3, CH3OH, 125°C

Rh/C(5%),

NH3,H2O,75-110oC

Pt/C(5%),

NH3, H20, 125°C

Fig. 4-1 : Selectivities of butyronitrile hydrogénation using different catalysts,

conditions and solvents [5].

• Hydrogénation in acid solution: The produced amines are converted

into salts and thereby deactivated for a further reaction to higher

substituted amines.

• Hydrogénation under acylating conditions: In reactions with acetic

anhydride or methyl formate the primary amines are converted into

amides, which subsequently can be hydrolysed.

• Hydrogénation in presence of ammonia: Reactions that produce

secondary and tertiary amines by the cleavage of ammonia are inhibited

in the presence of ammonia.

In industrial nitrile hydrogénation the addition of ammonia is the method of

choice, which has no great disadvantages and produces considerable higher

selectivities towards the primary amine. A new, industrially applicable

method to increase the selectivity to primary amines is to modify the

catalyst by formaldehyde (see Chapter 4.4.3).

Page 14

4.1.3 Hydrogénation of nitriles to aldehydes

The hydrogénation of nitriles produces aldimines as intermediates.

Therefore, in principle it is possible to synthesize aldehydes via the partial

hydrogénation of nitriles and the subsequent hydrolysis of the produced

aldimines. This conversion is only successful with y- and 8-hydroxynitriles,

because the cyclisation of the corresponding aldimines to the saturated

2-aminotetrahydrofuranes and -pyranes is faster than the further

hydrogénation (Figure 4-2) [7, 8].

ÇN HÇ^NH H2N HO

CH CH(ÇH2)n (ÇH2)n

Pd,H2

-OH -*--*-

(CH2)n Ö H?0 (CH2)n' O

-OH -*-

R R HR HR

n = 2-3

Fig. 4-2: Hydrogénation and cyclisation of a hydroxynitrile [6, 8].

Based on these fundamentals, in sugar chemistry a process for chain

prolongation was developed. Hydrogen cyanide is added and the

intermediate is partially hydrogenated and subsequently hydrolysed.

The conditions for an intramolecular stabilisation do not exist in most

aliphatic and aromatic aldimines. Nevertheless, the nitrile hydrogénation

can be stopped after the uptake of one mole equivalent hydrogen if suitable

basic compounds are added and the secondary imine is produced. In

general, the more stable the formed secondary imines are, the higher is the

conversion.

Aldehydes can be produced from the secondary imines according to two

processes:

• Hydrolysis: Short heating with diluted mineral acids in alcohol or acetic

acid. The azomethines are split into aldehydes and the salts of the

released amines.

Page 15

• Displacement: Aldehydes which produce more stable secondary imines

displace the other aldehydes.

Another possibility, the direct hydrogénation of nitriles to aldehydes in

acidic medium, was investigated by Möltgen and Tinapp [9]. High

selectivities were obtained with Raney nickel as catalyst in acid solutions.

4.1.4 Hydrogénation of nitriles to hydrocarbons

With a mixed catalyst of nickel- and copper(II)-oxide (3:2) on silicic acid,

nitriles can be converted to hydrocarbons, e.g. p-aminobenzonitrile to

p-toluidine (80%) or cinnamic acid nitrile to propylbenzol (90%) [4, 6]. For

this conversion, the catalyst is prehydrogenated at 300°C and subsequently

a mixture of nitrile and hydrogen is added at similar temperatures. This

transformation is also possible with a molybdenum sulphide catalyst [6].

4.1.5 Hydrogénation and cyclisation

Cyclisations occur in catalytic hydrogénations if a functional group is

produced that can react with a second functional group located in suitable

distance. One of the best known examples is the reduction of y-hydroxyl-

acids to y-lactones. Schiffbases or primary amines in molecules containing a

second interacting group, such as ketones, esters, acids, amides, olefins and

various heterocyclic rings can be hydrogenated into cyclic amines [2, 3]. A

second nitrogen group can also lead to cyclic amines via the hydrogenolysis

of ammonia. Halogen atoms in the y- or 8-position favour cyclisations to

pyrrolines or piperidines.

Cyclisations are not limited to catalytic hydrogénation, but can also occur

during organic reductions. Because the favourable conditions for

cyclisations are not identical to the conditions for hydrogénation, the

cyclisation sometimes occurs when the reaction mixture is recovered. In

some cases, the cyclisation can be either suppressed or enforced in presence

of large quantities of ammonia or acylating solvent, respectively [2, 3, 6].

Page 16

4.2 Mechanistic considerations of nitrile hydrogénation

4.2.1 Historic development

In the literature results are generally discussed according to a mechanism

proposed by Braun et al. [10] in 1923, based on a competition between

heterogeneous hydrogénations and homogeneous condensations

(Figure 4-3, Figure 4-4).

H2 H2

R- EN :NH

RR NH2

Fig. 4-3: Hydrogénation of the nitrile, producing the intermediate imine which is

further hydrogenated to the primary amine.

With his experiments Braun et al. [10] could exclude the reaction pathway

to the secondary amine via the aldehyde intermediate formed by the reaction

of aldimine with water.

H9:NH-

RCH2NH2

NH?H2

NH,

R R

route A

NH,

,^\,

H2

N^ R R

route B

Fig. 4-4: Mechanism proposed by Braun et al. [10] for the reaction towards the

secondary amine.

Page 17

Two possible intermediates (Figure 4-4) were proposed, the 1-amino-

amine (route A) and the secondary imine (route B). After the reaction of the

nitrile with one mole equivalent of hydrogen, the aldimine is produced,

which can then react with hydrogen to produce a primary amine. As soon as

both the aldimine and the primary amine are present, they react to form the

Schiff'base by condensation to the 1-aminoamine and splitting of ammonia.

The SchiffbasQ can then be hydrogenated to the secondary amine (route B).

Alternatively, the amine and the aldimine can react to form the

1-aminoamine, which can be hydrogenolysed to the secondary amine (route

A).

In 1967, Greenfield [11] presented a similar scheme for the formation of

the tertiary amine (Figure 4-5). The aldimine reacts with the secondary

amine to produce the 1-aminotrialkylamine, which is hydrogenolysed to the

tertiary amine (route A). Greenfield also proposed the alternative reaction

path via the intermediate of the enamine (route B).

y-NH(CH2CH2R)2 R

=NH «*~

Fig. 4-5: Mechanism for the tertiary amine formation proposed by Greenfield [11].

In 1986, VolfandPasek [12] summarised the hydrogénation in a scheme

and pointed out that there exist two types of reactions: Typical hydrogéna¬

tions (A, B, E and H) and acid-base catalysed condensations (C, D, F and G,

Figure 4-6).

Page 18

R NH2

B H,

NH2

RCHoNH,-=N :5=t D/^ ^=^

NH2

R N

NH,

-NH,

/^KR ^NH r

R N f<nR N

HN(CH2R)2

,R "2

R Ni/

-R N

Fig. 4-6: Scheme for the hydrogénation process proposed by VolfandPasek [12].

In 1992, Dallons et al. [13] postulated a mechanism for the formation of

by-products in which a semi-hydrogenated intermediate adsorbed on the

catalyst reacts with a primary or a secondary amine. The resulting 1-amino-

alkylamine or 1-aminodialkylamine reacts further to the secondary imine or

the tertiary enamine (Figure 4-7).

R1

H

H9EN

^R

^N

NHR2CH2R1« R1

M]

NH2

NR2

-R1

M]

Fig. 4-7: Intermediates, chemisorbed on a metal of the catalyst, that lead to secondary

(R2=H) or tertiary amines (R2=CH2R1) in a mechanism proposed by Dallons

et al. [13].

Non-hydrogenating active sites (acid sites) are responsible for the

adsorption of the amine on the catalyst. Obviously, the support has a great

Page 19

influence on the selectivity. The adsorption of primary and secondary

amines as well as ammonia is enabled by the acid sites. Dallons et al.

proposed that acid sites on the support give a high selectivity in favour of

the primary amines because these amines are adsorbed on the sites that are

not neighbouring the hydrogénation sites. The formation of by-products is

therefore suppressed.

A fundamentally different mechanism was formulated by Verhaak et al.

[14-16] {Figure 4-8). The hydrogénation of acetonitrile in the gas phase was

investigated using various acid nickel catalysts. The acidity of the catalysts

was successfully decreased by modifying the reaction temperature and by

the addition of potassium as a promoter to achieve a higher selectivity

towards the primary amine. The quantity and strengths of the acid sites were

determined by temperature programmed desorption (TPD) of ammonia, and

a linear correlation was found between the selectivity and the quantity of the

acid sites.

In this mechanism, the imine produced during the hydrogénation can

either be further hydrogenated to a primary amine or desorb from the

catalyst (Figure 4-8). If the imine readsorbs on acid sites, the acid-catalysed

gas phase metallic function

H,

= R- :NH

^ R CH2—NH2

migration(gas phase)

-R CH2—N CH2—R*

acidic function

H+

R CH—NH2

^ RCH9—NH,

i2—i\in2

NH9

R CH2—NH2-CH—R

NH,

^ R CH2—N^=CH—R

Fig. 4-8: Mechanism proposed by Verhaak et al. [14-16] in which side reactions to

secondary and tertiary amines are catalysed by acid sites on the catalyst.

Page 20

side reactions occur (Figure 4-9) forming azomethine and enamine. These

substances then migrate back to the hydrogénation sites through the gas

phase and are hydrogenated to secondary or tertiary amines.

NH

©H

© NH,NH2

II

NH2

NH2

R'

NH2

©NH,

R'

-H

NH2

NH

R'

©

NH3

N

©-H -NH,

Vc/R

Fig. 4-9: Condensation reactions on the acid sites of the catalyst: I Chemisorbed

primary imine on the acid sites and its resonance structures, II reactions of the

primary imine to produce the secondary imine or the 1-aminoamine [16].

Huang andSachtler [17-22] presented a mechanism based on deuterium

exchange experiments. Acetonitrile was hydrogenated with D2 in the

presence of a ruthenium catalyst to produce CHD2CN and CH3CD2NH2

(Eq. 4.1). Deuterated acetonitrile (CD3CN) was hydrogenated with H2 to

produce CD3CH2ND2 and CDH2CN (Eq. 4.2).

2 CH3CN + 2 D2 CHD2CN + CH3CD2NH2 Eq. 4.1

Page 21

2 CD3CN + 2 H2 CDH2CN + CD3CH2ND2 Eq. 4.2

Based on these experiments, Huang and Sachtler formulated a

mechanism without an imine intermediate but instead with the formation of

intermediates that are chemisorbed by double bonds on the catalyst

(Figure 4-10). According to Huang et al. [18] and Rode et al. [23], acid sites

CN

+

CH3

CD2

N

Ru

CH3

„CN CD, CN

|+

CH3

CD2

1 ^CH3

CH2, NH

\/Ru

CH

Ru

NH2

Fig. 4-10: Mechanism proposed by Huang and Sachtler that explains the H2/D2 distri¬

bution observed in the experiments [22].

on the catalyst surface have no influence on the selectivity towards the

primary amine. This conclusion is based on two ideas [18]:

Secondary and tertiary amines are also formed on catalysts on neutral

supports.

Acid sites on a catalyst are neutralised in strongly basic media.

Coq et al. [24] postulated a mechanism (Figure 4-11) which includes

the side reactions and does not exclude Huang and Sachtler's [22]

mechanism. The mechanism of Verhaak et al. [14] is also not excluded. The

side reactions that lead to the secondary and tertiary amines can take place

on acid sites or on hydrogénation sites of the catalyst. A mechanism for the

condensations on the hydrogénation sites was postulated (Figure 4-11), in

which a primary amine reacts with the chemisorbed carbene E, or

alternatively the chemisorbed intermediates D and E condense.

Page 22

CH,

CH^=N

H

R r. N -r. n

* *

A

^" B

N

**

C2^.2

2H

CH2 NH2

* *

D

R NH2

**

E

^

Fig. 4-11 : Mechanism for the hydrogénation proposed by Coq et al. [24].

4.2.2 Models for the formation of side products

Nucleophilic attack on the imine carbon: Addition of water to the double

bond of the imine, forming an aldehyde via the aminole intermediate

(hydrolysis of the imine). This aldehyde then condenses with amines to

secondary imines that are hydrogenated to the corresponding amines

(Figure 4-12) [25, 26].

R CH^NHH,0

OH

R CH NH2

-NH, /O

R Cv

H,N^ ^R

H,

R NHCH^N,

OH

VH2° R C H

HN.

Fig. 4-12: Formation of by-products: Side reaction with water, forming the secondary

amine [26].

Page 23

Nucleophilic attack on the cyano carbon: Nucleophiles (water, alcohols

and amines) can attack the nitrile group (Figure 4-13).

H

I ,0H20 | /s

***"

[M] N •<*" [M]-—NH2—(5'

^C OH R

R

H

HOR' |[M] N^=C R <

*"[M] N

^C OR'

R

H

I /NR'NH,R' | //

<»

[M] N «» [M]~--NH2—C.

NC NHR' R

R

Fig. 4-13: Formation of by-products: Nucleophilic attack on the cyano carbon [26].

Insertion reactions: Insertion of the nitrile on chemisorbed

intermediates and by-products (Figure 4-14).

Electron transfer and C-C coupling reaction: One electron transfers

from the nitrile to the metallic catalyst, forming an iminoradical that can

dimerize (Figure 4-15).

Obviously, the reactions to hydrocarbons (at high temperatures) and the

cyclisation reactions (Chapter 4.1.4 and Chapter 4.1.5) can also diminish

the yields of primary amines.

Page 24

H NCR' /R[M] N CH2R < *"- [M] N^=C

R'

NCR' /[M] R <

"[M] C.

N CH2RH

N R

/ NCR' /[M]^C „

*•[M]^C X

R N=

NCR'

R

R'

[M]—R « [M]—N=C^R

Fig. 4-14: Formation of by-products: Insertion reactions [26].

R

C^=N [M]n

[M]n N^^C R m" [M]*""1'—N^=C*

<" [M]n N^=C

R R

Fig. 4-15: Formation of by-products: One electron transfer and C-C coupling reaction

[26].

4.2.3 Influence of reaction parameters on nitrile hydrogénation

Influence of ammonia

The effect of ammonia on the selectivity was first investigated by Braun et

al. [10]. All authors describe a higher selectivity towards the primary amine

if ammonia is added to the reaction mixture, but the reason of this effect is

unknown. Another unexplained effect is that the reaction rates are higher if

ammonia is added up to a certain ammonia concentration and are then again

lowered [27]. Several explanations were put forward:

Page 25

• Ammonia influences the equilibrium between the amine, the imine and

the azomethine (Figure 4-16) [28].

CH2 CH -NH3 ^. .CH

R NH2 R-^ ^nh **R N^ R

Fig. 4-16: Equilibrium of amine, imine, azomethine and ammonia [28].

• Ammonia reacts with the primary imine to produce the 1 -aminoamine,

which is hydrogenolysed to form the primary amine {Figure 4-17) [29].

<ZH NH3R ^NH -•

NH2

H, CH,- »- / ^

~* R NH2

NH,

Fig. 4-17: Addition of ammonia to the double bond of the imine and hydrogenolysis of

the 1-aminoamine [29].

Ammonia poisons the acid sites of the catalyst, inhibiting the acid-

catalysed side reactions [16].

Ammonia modifies the electronic properties of the catalyst, preventing

the unwanted side reactions.

A similar positive effect on the selectivity can be observed if alkali

hydroxides are added to the reaction mixture [30, 31].

Influence of the solvent

Besson et al. [32] investigated the effect of various solvents on the

selectivity and the activity by comparing the polarity of the solvent with the

selectivity towards the primary amine. It was found that the more polar the

solvent, the higher the selectivity. In addition, the activity in different

alcohols was investigated, and it was found that the reaction rate increased

Page 26

as the number of C-atoms in the alcohol increased. The solubility of

hydrogen in the solvent also influences the hydrogénation reaction.

Influence of the temperature

Generally, the selectivity decreases as the temperature increases [26].

Degischer [27] found that within the range 60-140°C the selectivity

increased linearly with increasing temperature. At temperatures above

160°C, the selectivity decreased with increasing temperature. This change

may be due to different activation energies at higher temperatures (kinetic

control) or a shift towards equilibrium at higher temperatures

(thermodynamic control).

Influence of the hydrogen pressure

A higher hydrogen pressure causes a higher reaction rate. Furthermore, a

higher pressure, in general, leads also to a higher selectivity. However, if

Raney nickel is employed, a higher hydrogen pressure surprisingly yields a

lower selectivity towards the primary amine [26].

Influence of the catalyst

The main influence on selectivity and activity is based on the metal catalyst

and its support. Often similar selectivities and activities are observed in the

liquid and the gas phase hydrogénation.

4.3 Reversible reactions of amines

The synthesis of secondary and tertiary amines starting from primary

amines was described by Nicodemus and Schmidt [33] in 1930. Ethylamine

and butylamine reacted at temperatures of about 220°C to the

corresponding secondary amines using a catalyst produced by coating

cobalt carbonate onto pumice stone. Selectivities of 76-78% towards the

secondary amine were observed at conversions of 65-70%. 1936 Herold and

Smykal [34] reported on the preparation of primary amines from secondary

Page 27

and tertiary amines and ammonia. At temperatures of 300-450°C, using an

excess of ammonia and catalysts such as alumina gel, activated carbon,

aluminium oxide, reduced nickel catalysts, yields up to 66% were achieved.

A commercially interesting application is the production of hexamethylene-

diamine starting from azepane, that was patented by Reppe and Bauer [35]

(Figure 4-18). In the presence of hydrogen, ammonia and a nickel or cobalt

Fig. 4-18: Production of hexamethylenediamine starting from azepane (I) and the

corresponding secondary amine (II).

catalyst azepane is converted to hexamethylenediamine "in good yields" in

the liquid phase at temperatures of 140-220°C. This reaction is of interest

because azepane is produced as a by-product during the hydrogénation of

adiponitrile. Also, the production of hexamethylenediamine from the

corresponding dimer and ammonia is an interesting reaction since the dimer

is another by-product of the adiponitrile hydrogénation (Figure 4-18).

The reverse reaction, the cyclisation of hexamethylenediamine, was

also investigated [36], and azepane was obtained at conversions of 84% in

the gas phase at temperatures of 350-380°C using chromium and vanadium

oxide catalysts. Nowadays, the production of secondary amines starting

with primary amines is one of the standard methods in the laboratory

[37-40].

Page 28

In 1993/94 Verhaak et al. [14-16] proposed a mechanism for the

hydrogénation of amines. Their attention focussed on the acid sites on the

catalyst and their role in the production of higher substituted amines as

by-products (vide Chapter 4.2.1, Figure 4-8 and Figure 4-9).

In another publication, the disproportionation of the propylamine in the

gas phase using a continuous flow reactor, hydrogen and hydrogénation

catalysts was investigated [41]. The production rate of dipropylamine

formation decreased with increasing hydrogen pressure. If the reaction was

run without hydrogen, the conversion to the secondary amine was decreased

and dipropylimine as well as propylimine were obtained as main products

(Figure 4-19). The reaction rates depend on the acid sites on the catalyst.

Fig. 4-19: Production of dipropylamine B (in the presence of hydrogen), or dipropyl¬

imine C and propylimine D (in the absence of hydrogen) [41].

A mechanism was proposed in which the disproportionation is divided

into four reaction steps (Figure 4-20): a dehydrogenation followed by an

acid catalysed condensation and finally the hydrogenolysis of the amino-

amine produced.

This reaction mechanism is supported by the fact that the number of the

acid sites determined by temperature programmed desorption of ammonia

Page 29

correlates with the conversion of propylamine in the disproportionation

experiments (Figure 4-8).

-NH3

Fig. 4-20: Reaction sequence for the disproportionation of primary amines proposed by

Verhaaketal. [41].

4.4 Raney nickel

4.4.1 Preparation methods

Since the work of Sabatier [42] nickel is known as a good hydrogénation

catalyst. To enlarge the reaction surface, the metal was dispersed on

inorganic supports. Another way to increase the activity was discovered by

Raney [43] in 1925. His patent describes a process to remove Si from a

NiSi-alloy with alkaline solutions. NiAl intermetallic components showed a

higher activity than those of Si [44]. The investigation of Raney nickel is

interesting due to its complex skeletal structure and its wide range of

application in organic synthesis. Raney nickel is often used in industry, e.g.,

in the catalytic hydrogénation of adiponitrile to hexamethylenediamine, or

in the hydrogénation of benzene to cyclohexane.

Page 30

Raney nickel is prepared by leaching the Al in a NiAl-alloy with a

sodium or potassium hydroxide solution in accordance with Eq. 4.3 [45,

46]. If the alkaline hydroxide is not used in large excess (20-30% NaOH or

30-40% KOH), the aluminate formed is deposited as bayerite on the catalyst

(Eq. 4.4).

2 Al + 2 OH" + 2 H20 ^=^ 2 A102" + 3 H2 Eq. 4.3

2 A102" + 4 H20 =5=*= A1203 x 3H20 + 20H_ Eq. 4.4

There are three methods that can be used to manufacture Raney nickel. In

the first method, the nickel particles are slowly added to an alkaline

solution. In the second method, the solution is slowly added to the alloy in a

neutral suspension. In both cases, it is important that the reaction is

controlled. If aluminium has to be removed quantitatively, the nickel

particles must be added slowly to an alkaline solution. When no further

hydrogen is evolved, the reaction mixture is heated in concentrated alkali

solution.

The concentration of the leach decreases with conversion, so that the

suspension must be decanted several times, and the lye replaced. The fresh

Raney nickel is stored in a 1 M sodium hydroxide solution. To minimize

variations in the properties of the catalyst, all samples should be taken from

the same batch.

4.4.2 Industrial applications

Raney nickel is a heterogeneous catalyst often used, with many applications

in hydrogénation reactions [47-49], especially in the hydrogénation of

nitriles to primary amines which are used in polymeristion reactions. This is

due to the good selectivity and the cheap price of the metal compared to

other metals used in hydrogénation processes. Another field of application

is the hydrogénation of aromatic nitro compounds [50], of C-C double

bonds and even as a substitute for the Lindlar catalyst (Table 4-1).

Page 31

Table 4-1 : Examples for the applications ofRaney nickel in industrial processes [47-49].

reaction substrate product application

functional group

hydrogénation 2,4-dinitrotoluene 2,4-diaminotoluene polyurethanes

nitro

hydrogénation 1,5,9-cyclododeca- cyclododecane polyesters from

diene triene nylon-6,12

hydrogénation C10-C13-3-ketoacid Cio"Ci3-3- pharma products

ketone hydroxyacid

hydrogénation 2-ethylhexanal 2-ethylhexanol plasticiser

aldehyde

hydrogénation stearonitrile stearylamine plasticiser

nitrile

hydrogénation adiponitrile hexamethylenedi - nylon-6,6

dinitrile amine

hydrogénation 1,4-butynediol 1,4-butanediol THF

alkyne

hydrogénation benzene cyclohexene polyamides

aromatic

hydrogénation phenol cyclohexanol polyamides

aromatic

aminolysis 1,6-hexanediol hexamethylenedi - nylon-6,6

alcohol amine

alkylation dodecylamine dimethyldodecyl- surfactants

amine amine

4.4.3 Variation of the properties ofRaney nickel

Various methods were tested to obtain more selective catalysts for

different processes, mainly in the hydrogénation of nitriles and of aromatic

nitro compounds. There, the main parameters are the metal composition

Page 32

(ratio of nickel/aluminium) [50-52], the strength of the basic treatment, the

doping with other metals [53-55], the process parameters during the

production process (e.g. quenching the parent alloy in cold water [56-58])

and the modifying additives such as lithium hydroxide [31], morpholine

[59], copper acetate [60], formamidine salts [61] or vanadium salts [62]. A

recently published patent of Degischer and Rössler [1] is presenting the

advantages of a catalyst modified by formaldehyde in the hydrogénation of

nitriles. A yield of 96.4% was obtained if 'pynitrile' (4-amino-5-cyano-2-

methoxypyrimidine) was hydrogenated using a commercially obtainable

Raney nickel. The selectivity was increased to 99.6% if the catalyst was

treated with a 1% formaldehyde solution prior to the hydrogénation

reaction. Similar effects on the selectivity were found if instead of

formaldehyde carbon monoxide (98.8% primary amine) or acetaldehyde

(97.3% primary amine) were used to modify the catalyst [63]. Further

experiments with benzonitrile as a model substance are presented in

Figure 4-21. The disadvantage of such a modification is the loss of activity,

which is leading to a higher catalyst load in industrial processes [64].

In addition, experiments with Raney cobalt and nickel-on-carrier

catalysts were patented. Again, the modification by formaldehyde has a

positive effect on the selectivity towards the primary amine. The selectivity

for the primary amine increased from 96.8% to 98.1% if a nickel-on-carrier

was modified using formaldehyde, and from 96.8% to 99.1% if a Raney

cobalt catalyst was modified.

4.4.4 Adsorption of nitriles and their hydrogénation intermediates

Recently, great efforts have been made to describe the adsorption of

hydrogen as well as of other substances involved in the catalytic

hydrogénation of nitriles. Semi-empirical studies, molecular modellings,

studies on metalorganic substances as well as spectroscopic data (high

resolution electron energy loss vibrational spectroscopy: HREELS) were

reported [26, 65-69]. De Bellefon and Fouilloux [26] summarised the

chemisorbed species in a scheme as shown in Figure 4-22.

Page 33

ww

CO

E

0

g

Ero

>^N£Z

0

100

no modification

1%CH20 modified

2.5% CH20 modified

5% ChLO modified

120 140 160 180 200

reaction time / [min]

Fig. 4-21 : Hydrogénation of benzonitrile using a differently modified catalyst. Reaction

conditions: 100 ml benzonitrile, 30 g ammonia, 2 1 methanol, 5.7 g Raney

nickel, 100°C, 4 MPa and 1200 rpm [1, 64].

Because not only the knowledge of the interaction between nitriles and

their hydrogénation intermediates with the catalyst is important to

understand the hydrogénation, Blyholder and Neff [70] investigated the

adsorption properties of solvents, such as methanol, ethanol, diethyl ether

and water using a nickel-on-siliciumoxide catalyst. The adsorbed species

were observed with infrared spectroscopy. It was found that methanol reacts

at a temperature of 20°C and produces chemisorbed CO on the catalyst

surface. Ethanol reacts in the same way, so that in addition to the infrared

band of CO also bands of CH3 and CH2 were detected. Ni-CH3,

Ni-CH2-CH3 and Ni-0-CH2-CH3 are supposed to be chemisorbed at the

surface. Neither water nor diethyl ether chemisorbed or reacted with the

surface.

From these experiments Blyholder and Neff'[70] concluded that carbon

monoxide is the only species that is chemisorbed on nickel surfaces. The

Page 34

RCN RCN* RCN* + H RCN* + 2H RCN* + 3H RCN* + 4H RCN + 4H

R

C

R H R R

CH2 11 HCs

R

CH2

R

CH2

y y N

*1 ul N\

N XNH NH NH2

/ [M] [M] [M] [M] [M] [M]

' end-on nitrile metaloimrne amido end-on imine amino amine

R—CH=^NH R -CH2 NH2

free nitrile

[M]

side-on nitrile

\

[M]

side-on inline

free amine

\

R\C^NH R\ /NHl

H

R f ^NH2

[M] [M] [M]

lminoacyl aminocarbene aminoalkyl

Fig. 4-22: Catalytic hydrogénation of nitriles: Intermediates adsorbed on a metal centre

of the catalyst [M] [26].

reaction pathway of alcohols leading to CO produces aldehyde or ketone as

intermediates. Initially, the alcohol is dehydrogenated before the C-C and

the C-H bonds are broken and CO and other fragments are produced.

Diethyl ether does not react with nickel surfaces as it can not dehydrogenate

a hydroxyl group. The dehydrogenation of alcohols on Raney nickel was

also reported by Besson et al. [25].

4.4.5 Inhibition/poisoning of the catalyst

Every process, chemical or physical, that reduces the activity of a catalyst

can be classified as deactivation. In most cases, especially in complex

reactions or when using complex catalysts, a change in activity is

accompanied by changes in selectivity. A change in selectivity can be

achieved using poisons or inhibitors. The following processes are important

[71,72]:

Irreversible adsorption of an inhibitor at the surface or reaction of an

inhibitor with the catalyst surface.

Page 35

Competitive reversible adsorption of a poison on the surface.

Poison induced restructuring of the surface.

• Physical or chemical blocking of the pores.

With the example of a hydrocracking reaction, Penchev et al. [73]

demonstrated that the selectivity can be changed using different poisons

(thiophene poisons the metallic sites, pyridine the acid sites).

In the case of Raney nickel acetonitrile desorbs at temperatures above

75°C, while more strongly adsorbed nitrile fragments or molecules desorb

at temperatures above 180°C. The decomposition of acetonitrile forms two

carbon compounds that deactivate the catalyst. McCarty and Wise [74] and

Kock et al. [75] identified the adsorbed species as oc-carbon and nickel

carbide. These compounds can be hydrogenated at temperatures above

200°C, thereby restoring the catalyst's activity. Thus, the deactivation of the

catalyst during the hydrogénation of acetonitrile can be prevented by

applying high hydrogen pressures [24, 76].

4.4.6 Acid sites

The influence of acid sites on the selectivity of the hydrogénation of nitriles

is still disputed. Verhaak et al. [14-16] described the influence of acid sites

present during the hydrogénation of nitriles on the selectivity and the cata¬

lyst activity using nickel on a silicon and magnesium support. These acid

sites are formed during the synthesis of the catalyst and are ascribed to the

support. Their amount and strength were determined by the temperature

programmed desorption (TPD) of ammonia.

Increasing the number of acid sites leads to higher reaction rates and a

decreased selectivity. These observations were confirmed by Cabello et al.

[28] and Dung et al. [77].

Other authors [12, 17-22] did not confirm the influence of acid sites and

point out that the selectivity is influenced only by the metal used and by the

reaction conditions.

Aluminiumoxide is one of the most commonly used support materials in

heterogeneous catalysis. Examples are Pt-Re/Al203 (reforming) and

Co-Mo/Al203 (dehydrosulfonation). This support is thermally stable and

Page 36

allows an appropriate distribution of catalytically active compounds. How¬

ever, as aluminiumoxide is not an inert support, many reactions are cataly¬

sed by it, e.g. double bond migrations, E/Z isomerisations of olefins and

H/D exchange in hydrocarbons.

Knözinger [78] determined the acid sites via the infrared adsorption of

CO and measuring the stretching vibrations of C-0 and O-H bonds, which

depend on the bond strength (Figure 4-23). CO is also chemisorbed by

Lewis acid sites via G-bond and 7t-rebond. Knözinger determined the

oxidation state and the coordination number of cations on the surface and

the relative bond strengths of metal-CO bonds. In addition, acid site

sequences were determined for hydroxyl groups on the surfaces.

M—O—H— CO

Fig. 4-23: Chemisorption of CO on surface hydroxyl groups [78].

Page 37

Page 38

Chapter *J7

Hydrogénation of butyronitrile

5.1 Reaction system

The hydrogénation of butyronitrile leads to butylamine as main product if

Raney nickel is used as hydrogénation catalyst. A scheme of the reaction is

given in Figure 5-1. The hydrogénation experiments were made in a 500 ml

steel hydrogenator (see Chapter 9.1.1) according to the procedure described

in Chapter 9.2.1. The reversibility experiments presented in Chapter 5.6

were made in the 500 ml steel hydrogenator or in 35 ml screening

autoclaves (see Chapter 9.1.4) according to the procedure described in

Chapter 9.2.2. Two different catalysts were used, Raney nickel B 113 Z,

batch 20018989 from Degussa-Hillls AG and nickel-on-carrier Ni 1404 P,

lot H-99 form Engelhard.

Always when dibutylimine is mentioned, this means, that this is the

amount of dibutylimine measured by gas chromatographic analysis or

reaction rates extrapolated from gas chromatographic measured values.

However, one must keep in mind that it is possible, that other substances as

1-amino amines, amidines or imines react in the injector and the column of

the gas Chromatograph (at temperatures of 250°C) and then appear as

dibutylimine in the gas chromatographic spectrum.

Next to dibutylamine, tributylamine was observed as by-product.

Dibutylamide and the tertiary amidine were also formed in the reaction

mixture (Figure 5-2).

The product distribution of the hydrogénation is expressed in [mass-%],

the reaction rates in [mmol/s] or in [mmol/s*kg catalyst]. The indicated

pressure ascribes to the overall pressure, used during the reaction.

Page 39

CS

13 Q

C3

13 Q

+

4= »

Ö

J3

-&\43 P5

l-J-l ^t-

+ ^

<n

oN

43

-d

u

Ö

+

öo

43 P5

K

43

Fig. 5-1 : Hydrogénation of butyronitrile (BN) to the desired primary amine (BA) and

the secondary amine (DBA) over butylimine (not observed by GC) and

dibutylimine (DBI) as intermediate.

Page 40

tributylamine

TBA

NH

N-butylbutanamide, dibutylamideDamide

N,N-dibutylbutyramidineTamidine

Fig. 5-2: By-products of the hydrogénation of butyronitrile, observed by GC.

In order to follow through a well structured discussion of the influence

of various reaction parameters on the rate and the selectivity of the

hydrogénation of nitriles, first a simple semi-quantitative macro-kinetic

model is presented. The basis for the derivation of the kinetic equations is

the mechanism depicted in Figure 5-1. For the sake of simplicity it is

assumed that the rates of the individual hydrogénation steps are of the x-th

order with respect to the partial pressure of hydrogen. Furthermore, based

on the experimental data it can be assumed, that for the concentrations of

the intermediates BI, DBDH+ and DBD the steady-state approximation of

Bodenstein [79] can be applied.

The differential selectivity ratio (d[BA]/d[DBA]) is given by (Eq. 5.3):

d[BA]=

dt

d[DBA]

dt

k2P?H2)[BI]

k5P(H2)[DBI]

d[BA]_

k2 [BI]

d[DBA] k5[DBI]

Eq. 5.1

Eq. 5.2

Eq. 5.3

Page 41

The ratio [BI]/[DBI] can be evaluated from the following equations

(Eq. 5.5-Eq. 5.8):

^JP = k1[BN]PH2-[BI](k_1 + k2p^ + k3[BA][H+]) Eq.5.4

+k 3[DBDH+]

d[D^H ]= k3[BI][BA][H+] +k 4[DBI][H+]p(NH3) Eq. 5.5

-(k4 + k_3)[DBDH+]

As the protolytic side equilibrium between DBDH+ and DBD is assumed to

be much faster than all the other partial reaction steps, it can be neglected

for the formulation of the kinetic equations. From the equations (Eq. 5.4)

and (Eq. 5.5) the equations (Eq. 5.6) and (Eq. 5.7), respectively follow:

k^BNJp^ + k 3[DBDH+] -

ffiü

[BI] = — Eq. 5.6

k-l+k2P?H2) + k3([BA][H ])

k3[BI][BA][H+]+k4[DBI][H+]p(NH3)-d[D^H ]

[DBDH+] = —— - Eq. 5.7

K4 + K_3

If the steady-state approximation of Bodenstein [79] applies for the

intermediates BI, DBDH+ and DBD, i.e. if

d[BI]

dt«k1[BN]p^H2) + k_3[DBAD]

and

d[DBDH+] «k [BI][BA][H ]+k4[DBI][H ]p(NH

dt{ 3

Page 42

it follows:

_

k^tBNlp^j + MDBIHH Jp^ + kjkJBNJp^[B1J — Eq. 5.8

(k_3 + k4)(k_! + k2p^H2)) + k3k4[BA][H ]

and thus

[BI]_

1(k-3 + k4)k1[BN]p^)[5^] + k_3k_4[H+]p(NH3DBI] k4 (k 3

+ k4\ x +

1^J(k-1+k2P(H2))+ k3[BA][H]

Eq. 5.9

From the equations (Eq. 5.3) and (Eq. 5.9), the differential selectivity

(Eq. 5.10) follows:

d[BA] k2(k-3 + k4)k1[BN]P(H2)[5^ + k_3k_4[H+]p(NH3)Eq. j.lu

d[DBA] k4k5 As + k4^ x +

[-^J(k_1+k2p^H2))+ k3[BA][H ]

Equation (Eq. 5.10) reveals a very diversified and complex influence of the

selectivity on the various reaction parameters. A given reaction parameter,

such as e. g. the partial pressure of hydrogen, can either increase or decrease

the selectivity with respect to the desired butylamine, depending on the

relative rates of the individual reaction steps. This diversity shell be

exemplified by the following few cases:

Case I: k2p^2) » k_Y ; k_3 » k4 ; k_4[H+]p(NH3) » ki[BN]P*H2)i5^j

d[BA]_

k2k-3 k_4[R ]P(NH3d[DBA] k5 k_3k2p*H2) + k3k4[BA][H+]

Eq. 5.11

Case II: k2p*Ui) « k_Y ; k_3 » k4 ; k_4[H+]p(NH3) » ki[BN]P?H2)i5^7J

Page 43

d[BA]_

k2k_3 k_4[H ]p(NH3)e ^ ^

d[DBA] k5 k.gk^ + kgkJBAHH*]

Case III: k2p^2) » k_j ; k_3 « k4 ; k_3k_4[H+]p(NH3) » kik4[BN]P*H2)i5^7j

d[BA]_

k2k-3 k-4[H ]P(nh3)E 5 13

d[DBA] k4k5k2p^2) + k3[BA][H+]

Case IV: k2p*Ui) « k_Y ; k_3 « k4 ; k_3k_4[H+]p(NH3) » kik4[BN]P*H2)i5^7j

d[BA]_

k2k_3 k_4[H ]p(NHad[DBA] ^sk^+^CBA]^]

Case V: k2p^2) » k_Y ; k_3 « k4 ; k_4[H+]p(NH3) « ki[BN]P*H2)i5^j

d[BA] _k2k-3 kl[BN]p^)[DBli

Eq. 5.14

d[DBA] k5 k_3k2p^H2) + k3k4[BA][H+]Eq. 5.15

Case VI: k2p^2) « k_j ; k_3 » k4 ; k_4[H+]p(NH3) « ki[BN]P*H2)i5^j

d[BA]_

k2k_3 ki[BN]P?H2)[5Blid[DBA] k5 k^k^ + k^JBAHH^

Eq. 5.16

Page 44

Case VII: k2p*H2) » k_Y ; k_3 « k4 ; k_3k_4[H+]pNH3 « ^k^BN^2DBI

d[BA]_

k2 ki[BN]p^)ipiriEq 517

d[DBA] k5k2p^2) + k3[BA][H+]

Case VIII: k2p^2) « k_j ;k_3 « k4 ; k_3k_4[H+]p(NH3) « k4ki[BN]P*H2)i5^7j

d[BA]_

k2kl[BN]p^)î5ïïïiEq51g

d[DBA] k5k,+kJBA][H+]

5.2 Thermodynamic aspects

A profound investigation of the thermodynamic equilibrium of the

hydrogénation of butylamine within a temperature range of 25 to 150°C and

a hydrogen pressure range of 1 to 20 MPa partial pressure of hydrogen was

made by Chojecki [80] using HSC Chemistry software. The best

thermodynamic selectivity was calculated at high hydrogen pressures and

low temperatures (75% butylamine at 20 MPa hydrogen and 298 K). The

thermodynamic selectivity decreased dramatically at lower hydrogen

pressures, while temperature had only a small influence.

5.3 Aspects of mass and heat transport

The kinetics of multiphase reactions are not only influenced by the chemical

kinetic but also by the rate of mass and heat transfer. The investigated

system presented in this study is a three phase hydrogénation consisting of

gaseous hydrogen, liquid substrate and a solid catalyst. Mass transfer of

Page 45

hydrogen from the gaseous phase into the liquid phase as well as that of

hydrogen and the substrate (A) from the liquid phase to the surface of the

catalyst can significantly influence the kinetics of the individual reaction

steps, and thereby, also the selectivity of the overall hydrogénation process

(Figure 5-3).

catalyst S

Ca,s

Ch2,s

r

Fig. 5-3: Model of a three phase mass transport: Mass transport from the gaseous into

the liquid phase and from the liquid phase to the surface of the catalyst.

Cy: concentration of reagent i in phase j. i:H2 = hydrogen, iA = substrate A;

a,b,c: diffusional boundary layers.

Heat transfer from the surface of the catalyst to the liquid phase can also

influence the selectivity. This can occur if the reaction rates are much faster

than the heat transfer. In the case of a highly exothermic reaction such as the

hydrogénation of nitriles, the temperature on the surface of the catalyst is

higher than the measured temperature in the bulk liquid phase (Figure 5-4).

This, in turn, increases the local hydrogénation rate which than increases the

local temperature even further (run away). A different temperature effect on

C,, gas phase G

CH2,G

Page 46

Tu

liquid phase L

V

catalyst S

Fig. 5-4 : Temperature profile of an exothermic reaction: Heat transfer from the surface

of the catalyst to the liquid phase.

the rates of the competing reactions leads to a selectivity behavior which is

influenced by heat transfer.

5.4 Influence of reaction parameters on nitrile hydrogénation

The influence of the different reaction parameters on the initial differential

selectivity, the product distribution and the activity of the catalysts are

discussed in this Chapter. The product distribution (as yield butylamine after

full conversion) is plotted versus the initial differential selectivity

(calculated by dividing the initial production rate of butylamine d[BA]/dt by

the initial production rate of dibutylamine d[DBA]/dt) in Figure 5-5. It

seems quite evident that an interdependence of these parameters exists and

that with increasing values of the initial differential selectivity the yield of

butylamine approaches a maximum limiting value which is definitely lower

than 100%.

Page 47

92-

91 -

90-

,—,-

89-

CO -

CO

CD88-

F -

HC-*-~~~

Cl>"

c 86-

E -

CD 8b->s

3

.Q84-

Ti'

CD83-

>s-

82-

81 -

an-

parameter:stirrer speed

• mass catalystA temperatureV recycling

pressure

0 5 10 15 20 25 30 35 40 45 50

initial differential selectivity / [-]

Fig. 5-5: Initial differential selectivity versus product distribution (yield butylamine).

5.4.1 Influence of the overall pressure

The influence of the overall pressure on the product distribution, initial rates

and initial differential selectivity was investigated for 240 g BN, 3.75 g

RaNi, 100°C, and 1000 rpm in a range from 1 to 6 MPa. The results are

summarised in Figure 5-6, Figure 5-7 and Figure 5-8. The selectivity

towards the primary amine is decreasing with higher partial pressures of

hydrogen. Such a selectivity behavior is predicted by the approximate

equation (Eq. 5.10), mainly for the special cases I and III. The maximum

amount of the intermediate dibutylimine is almost independent of the

hydrogen pressure. The initial rates are rising using higher pressures.

The large scattering range of the measured initial reaction rates is due to

technical problems with the cascade regulation of the temperature. When

large amounts of hydrogen are pressed into the reactor, the temperature is

decreasing due to the cold gas. This is compensated by the temperature

regulation. If the reaction is then started, the reactor is heated from outside

as well as from inside by the exothermic reaction. The temperature

controlling cascade regulation is compensating this trend and starts to cool.

Page 48

au -

^

85-

-—B-^^B —

—m-ï£ 80-

CO

—— butylamine

—

CO

I75-• dibutylamine

tributylamine

o 70=.T- max. dibutylimine

-."

~V- T-

~vproductdistribut

O

Ol

O

.

I.I.I.-*- T-~T~ -

-

-

5-

0- 1 1 1 ' 1A A à. È. *

i '

2 3 4 5

overall pressure / [MPa]

Fig. 5-6: Influence of the overall pressure on the product distribution. Reaction

conditions: 240 g BN, 3.75 g RaNi, 100°C and 1000 rpm.

550-

500-

450-

400-

cd 350-h—»

cc

'Z. 300-

o

'5 250-

CD

£ 200-

CD

« 150-^g

100-

50-

0-

-d[BN]/dt / [mmol/(s*kg)]

• d[BA]/dt/[mmol/(s*kg)]* d[DBA]/dt/[E-1 mmol/(s*kg)]

d[DBI]/dt/[E-1 mmol/(s*kg)]


Fig. 5-7: Influence of the overall pressure on the initial reaction rates. Reaction

conditions: 240 g BN, 3.75 g RaNi, 100°C and 1000 rpm.

Page 49

28-

26-

24-

-22-1

I 20-

8 18-

CD

'£ 16-

CD

| 14-

1 12-

~

ID¬

S'


Fig. 5-8: Influence of the overall pressure on the initial differential selectivity.

Reaction conditions: 240 g BN, 3.75 g RaNi, 100°C and 1000 rpm.

However, due to a heat transfer delay the cooling is not fast enough. As a

consequence, a certain thermal runaway occurs. Thus, the difference

between the measured and the effective reaction temperature amounts in

some cases up to 10°C. This in turn has a massive effect on the initial rates.

5.4.2 Influence of the temperature

The influence of the temperature on the product distribution, the initial rates

and the initial differential selectivity was investigated in a range of 60-

120°C using standard conditions (240 g BN, 3.75 g RaNi, 1 MPa,

1000 rpm). The results are presented in Figure 5-9, Figure 5-10 and

Figure 5-11. The increase in selectivity towards butylamine can be

explained by the lower activation energy for the hydrogénation to dibutyl¬

amine. Activation energies were determined using a "quasi" Arrhenius plot

(Figure 5-12). The conditions of 240 g BN, 100°C, IMPa and 1000 rpm are

near the gas-liquid hydrogen transfer limitation so that the experiments (and

Page 50

95-1

90-

^85-

\ 80-

CD

Ê75-I

o 70'-4—'3

E 15 H

is iono

T3O

5-

0-

-5

butylamine

dibutylamine

max. dibutylimine

60 70 80 90 100 110 120

temperature / [°C]

Fig. 5-9 : Influence of the temperature on the product distribution. Reaction conditions:

240 g BN, 3.75 g RaNi, 1 MPa and 1000 rpm.

275-

250-

225-

200-co

CD

TO 175-

0 150-

1 125 H

ro 100-

^ 75-

50

25

0


• d[BA]/dt / [mmol/(s*kg)]* d[DBA]/dt/[E-1 mmol/(s*kg)]

d[DBI]/dt / [E-1 mmol/(s*kg)]

60 70 80 90 100 110 120

temperature / [°C]

Fig. 5-10: Influence of the temperature on the initial rates of butyronitrile

hydrogénation. Reaction conditions: 240 g BN, 3.75 g RaNi, 1 MPa and 1000

rpm

Page 51

35 n

30-

-25-1

'>

'sM 20-CDCO

"cd

'E 15-

CD

=5 10-

"CD

5-

—1 1 1 1 1 1 1 1 1 1 1 1-

60 70 80 90 100 110

—I '

120

temperature / [°C]

Fig. 5-11: Influence of the temperature on the initial differential selectivity. Reaction

conditions: 240 g BN, 3.75 g RaNi, 1 MPa and 1000 rpm.

6.0-1

5.5-

5.0-

4.5-

4.0-

3.5-

3.3.O-

^2.5-

2.0-

1.5-

1.0-

0.5-

0.0'

0.0025

In (d[BN]/dt / [mmol/(s*kg)])• In (d[BA]/dt / [mmol/(s*kg)])* In (d[DBA]/dt / [E-1 mmol/(s*kg)])

In (d[DBI]/dt / [E-1 mmol/(s*kg)])

—i—

0.0026

1

0.0027

1 1 r

0.0028

1/T/[1/K]

—I—

0.0029 0.0030

1

0.0031

Fig. 5-12: "Quasi" Arrhenius plot obtained from the initial reaction rates r0 (see

Figure 5-10).

Page 52

Table 5-1 : Activation energies determined from the "quasi" Arrhenius plot by linear fit.

rates r0 temperature range activation energy error (±)

/ [see Figure 5-12] /[°C] / [kJ/mol] / [kJ/mol]

-d[BN]/dt 60-100 48.9 2.2

-d[BN]/dt 100-120 26.9 1.8

d[BA]/dt 60-100 50.7 1.8

d[BA]/dt 100-120 23.8 2.7

d[DBI]/dt 60-80 63.4 2.5

d[DBI]/dt 80-120 37.9 0.6

d[DBA]/dt 60-80 39.5 0.9

d[DBA]/dt 80-120 15.8 2.2

specially the reaction rates) using temperatures above 100°C are mainly

influenced by this limitation. The ln(-d[BN]/dt) versus 1/T data show a kink

in the linear correlation at a value of 0.00268 (100°C). The linear fit in the

range of 60 to 100°C gives an activation energy of 49 kJ/mol, above 100 °C

an activation energy of about 27 kJ/mol. Interesting data points were

obtained for the production rates of dibutylimine and dibutylamine. A sharp

break at the linear plot at 0.00283 (corresponding to 80°C) can be observed.

An explanation for these sharp breaks in the linear plots can be given either

by a diffusion limitation of the reagents to the corresponding hydrogénation

or acid sites, respectively, or by a change of the rate determining steps (see

Chapter 5.3). As the activation energy is approximately half as large at high

temperatures than at low temperatures the explanation of a diffusion

limitation is more probable. A summary of the obtained activation energies

is given in Table 5-1.

Page 53

5.4.3 Influence of the ratio of catalyst to substrate

The influence of the catalyst to substrate ratio on the product distribution,

the initial rates and the initial differential selectivity was investigated to

exclude a gas-liquid transport limitation for hydrogen. The following

reaction parameters were chosen: 240 g BN, 100°C, 1 MPa and 1000 rpm.

The catalyst amount was varied from 1.25 g to 20 g (0.5-8.3% catalyst). The

results are plotted in Figure 5-13, Figure 5-14 and Figure 5-15. The

T ! { ! { ! { ^

0 5 10 15 20

mass catalyst / [g]

Fig. 5-13 : Product distribution using different amounts of catalyst. Reaction conditions:

240 g BN, 100°C, 1 MPa and 1000 rpm.

selectivity towards the primary amine is rising using larger amounts of

catalyst. This may be due to catalyst deactivation (see Chapter 5.4.4). The

initial reaction rates are rising linearly with the catalyst amount until 5 g

catalyst. This is an indication that the hydrogen concentration in the liquid

phase is constant and no mass transfer limitation exists for hydrogen within

this range. Above a catalyst amount of 5 g the reaction rates approach a

Page 54

IDU -

140-()

„

— 120-"~-~.

>?'>

=o 10°-

cu

cuw

80-

.59.

c

2 60-CD

iffT3

1 40"

^->

C_

20-

0- 1 1 1 1 1 1 '

10 15

mass catalyst / [g]

20

Fig. 5-14: Initial differential selectivity using different amounts of catalyst. Reaction

conditions: 240 g BN, 100°C, 1 MPa and 1000 rpm.

1 6-

1 4-

1 2-

cu

Id 10-

o

=6 o8'

CD

çu"S O6'CD

Ë

04-

02-

00-

-— -d[BN]/dt / [mmol/s]-•— d[BA]/dt / [mmol/s]* d[DBA]/dt/[E-1 mmol/s]

-—d[DBI]/dt/[E-1 mmol/s]

10 15 20

mass catalyst / [g]

Fig. 5-15 : Initial reaction rates using different amounts of catalyst. Reaction conditions:

240 g BN, 100°C, 1 MPa and 1000 rpm.

Page 55

limiting value. This limit is given by the rate of hydrogen transfer to the

catalyst surface (see Chapter 5.3).

An explanation for the increasing selectivity towards butylamine using

higher catalyst loadings can be twofold: With an increasing amount of

catalyst per reaction volume the hydrogénation rate increases and thus also

the local temperature due to the exothermic reaction. This in turn accelerates

the reaction even more (runaway) and thus decreases the local hydrogen

concentration at the catalyst surface. These local temperature and hydrogen

concentration gradients are the larger the higher the catalyst concentration

is. The stirring is not efficient enough to cancel out these gradients. As a

consequence, a larger amount of catalyst per reaction volume decreases the

local hydrogen concentration and increases the local temperature. Both

effects result in an increase in selectivity towards butylamine.

Such a selectivity behavior is predicted by the approximate equation

(Eq. 5.10), mainly for the special cases I and III.

If the stirrer speed is risen from 1000 rpm to 1500 rpm the reaction rates

rose only slightly whereas the selectivity did not change. These results are

shown in Table 5-2 (the conditions were: 240 g BN, 3.75 g RaNi, 100°C and

Table 5-2: Selectivity and reaction rates at different stirrer speed. Reaction conditions:

240 g BN, 3.75 g RaNi, 100°C and 1 MPa.

stirrer speed [BA] [DBA] [max. DBI] -d[BN]/dt d[BA]/dt

/ [rpm] / [mass-%] / [mass-%] / [mass-%] / [mmol/(s*kg)] / [mmol/(s*kg)]

1000 88.6 11.0 9.95 118 83

1500 88.7 10.6 10.6 131 87

1 MPa). This is a second indication that no gas liquid transfer limitation for

hydrogen exists at the following conditions: 240 g BN, 3.75 g RaNi, 100°C,

1 MPa and 1000 rpm. An important note is, that this conditions are near the

gas-liquid transfer limitation for hydrogen.

Page 56

5.4.4 Recycling of the catalyst

Experiments to investigate the catalyst deactivation were carried out at the

following conditions: 240 g BN, 3.75 g RaNi, 100°C, 1 MPa and 1000 rpm.

The results are presented in Figure 5-16 Figure 5-17, Figure 5-18. The

—— butylamine

—•—dibutylamine* tributylamidine

—— max. dibutylimine

12 3 4

cycle / [-]

Fig. 5-16: Product distribution of butyronitrile hydrogénation recycling the same

catalyst four times. Reaction conditions: 240 g BN, 3.75 g RaNi, 100°C, 1

MPa and 1000 rpm.

selectivity towards the primary amine as well as the activity and the initial

differential selectivity are decreasing. In both cases, the decrease is large

from the first to the second cycle and small in the following cycles. Due to

this decrease in activity it follows that the local gradients of temperature and

hydrogen concentration are becoming less and less pronounced (see

Chapter 5.3). Following what has been said in Chapter 5.4.3 about the

effects of these local gradients one must expect a decrease in selectivity with

increasing the catalyst cycles. Such a selectivity behavior is predicted by the

approximate equation (Eq. 5.10) mainly for the special cases II and IV.

Page 57

60-

50-

40-

0)

"cd

g"G

S 30-

i_

nj

c 20-

10-


d[BA]/dt / [mmol/(s*kg)]

d[DBA]/dt/[E-1 mmol/(s*kg)]


cycle / [-]

Fig. 5-17: Initial rates of butyronitrile hydrogénation recycling the same catalysts four

times. Reaction conditions: 240 g BN, 3.75 g RaNi, 100°C, 1 MPa and

1000 rpm.

18-1

16-

14-1

^ .

>s

>12-

ocu

CD10-

CO.

CD8-

c

fl) .

l_

ë R-

T3 -

CD 4-

C-

2-

0-

cycle / [-]

Fig. 5-18 : Initial differential selectivity recycling the same catalysts four times. Reaction

conditions: 240 g BN, 3.75 g RaNi, 100°C, 1 MPa and 1000 rpm.

Page 58

5.4.5 Influence of additives

The influence of additives on the selectivity and activity was investigated

using standard conditions (240 g BN, 3.75 g RaNi, 100°C, 1 MPa, 1000

rpm). The results are presented in Figure 5-19 and Figure 5-20. The

100-

95-

II butylamine

liiilll dibutylamine

H tributylamidine

I 1 max. dibutylimine

additive

Fig. 5-19: Influence of additives on the product distribution. Reaction conditions: 240 g

BN, 3.75 g RaNi, 100 °C, 1 MPa and 1000 rpm.

following additives were tested: EDS, pyridine and ammonia. If EDS is

used, the reaction rate as well as the selectivity towards butylamine are

decreasing. EDS is known as a poison for hydrogénation catalysts. An

explanation for this results is, that EDS is poisoning the hydrogénation sites

of the catalyst and thereby slowing down the hydrogénation (-d[BN]/dt and

d[BA]/dt). Because EDS does not poison the acid sites, the side reactions

(BI DBI) are not slowed down and the measured selectivity towards

butylamine is lowered. Addition of pyridine does neither change the

selectivity nor the reaction rates significantly. If ammonia is added, the

selectivity towards butylamine is increased. The hydrogénation reactions are

Page 59

220-1

200-

180-

160-

\ZZ\ -d[BN]/dt/[mmol/(s*kg)]

^M d[BA]/dt/[mmol/(s*kg)]

^M d[DBA]/dt/[E-1 mmol/(s*kg)]

i I d[DBI]/dt/[E-1 mmol/(s*kg)]

,«PS

additive

Z1*«V' ai.6^w' 9^X A^^,d^e

5<â^

Fig. 5-20: Influence of additives on the initial rates of butyronitrile hydrogénation. Re¬

action conditions: 240 g BN, 3.75 g RaNi, 100 °C, 1 MPa and 1000 rpm.

slowed down, but not as dramatically as the side reactions (BI DBI).

The maximum intermediate concentration of dibutylimine is reduced too.

Explanations for this behavior are given in Chapter 4.2.3.

5.5 Influence of washing/modification procedures

5.5.1 Influence of washing procedures with different solvents

In order to describe the changes of catalyst properties brought about by

various washing procedures, the washed catalysts were tested using the

butyronitrile hydrogénation reaction. Thereby, standard conditions were

used: 240 g BN, 3.75 g RaNi, 100°C, 1 MPa, 1000 rpm. The results are

plotted in Figure 5-21 and Figure 5-22. The water treatment does neither

change the selectivity towards the primary amine nor the reaction rates.

Page 60

105-

100-

^

95-

butylamine

dibutylamine

tributylamidine

max. dibutylimine

fresh 3xH20 3xMeOH 3xEtOH 5% formaldehyde

Fig. 5-21: Influence of the washing procedure on the product distribution. Reaction

conditions: 240 g BN, 3.75 g RaNi, 100 °C, 1 MPa and 1000 rpm.


d[BA]/dt/[mmol/(s*kg)]

d[DBA]/dt / [E-1 mmol/(s*kg)]



Fig. 5-22: Influence of the washing procedure on the initial rates of butyronitrile

hydrogénation. Reaction conditions: 240 g BN, 3.75 g RaNi, 100 °C, 1 MPa

and 1000 rpm.

Page 61

Washing the fresh catalyst with methanol or modify it using formaldehyde

is dramatically decreasing (by a factor of 10) the reaction rates. The

selectivity towards butylamine is also decreasing. This is not in agreement

with a published patent of Degischer and Rössler. An explanation for this

behavior is that at the given conditions, the hydrogénation rates are so slow,

that the formation of secondary amine from butylamine already present in

the reaction mixture is lowering the yield of butylamine.

The influence of the washing procedure was also investigated under

conditions of gas-liquid transfer limitation for hydrogen (compare

Chapter 5.4.3). The following conditions were used: 240 g BN, 15 g RaNi,

100°C, 1 MPa and lOOOrpm. The results are plotted in Figure 5-23 and

110

105

100

£ 95

E 85

1 80

-g 15o

^10

5

0

fresh 3 x H20 3 x MeOH 5% formaldehyde

Fig. 5-23: Influence of the modification on the product distribution at conditions of a

hydrogen transfer limitation. Reaction conditions: 240 g BN, 15 g RaNi, 100

°C, 1 MPa and 1000 rpm.

Figure 5-24. Again the water washing does not change the selectivity and

only a small change in the reaction rates was measured. The reaction rates

for the methanol washed and the formaldehyde modified catalyst decrease

Page 62

butylamine

dibutylamine

tributylamidine

dibutylamide

fresh 3xH20 1xH20,3xMeOH 5% formaldehyde

Fig. 5-24: Influence of the modification on the initial rates of butyronitrile

hydrogénation at conditions of a hydrogen transfer limitation. Reaction

conditions: 240 g BN, 15 g RaNi, 100 °C, 1 MPa and 1000 rpm.

again, but the decrease is not as dramatically as if no hydrogen transfer

limitation exists, or in other words, the reaction rates using the unmodified

catalyst are slowed down because of the hydrogen transfer limitation. As

expected, the selectivity towards the primary amine is rising due to the local

hydrogen concentration (see Chapter 5.4.1) if high amounts of catalyst are

used.

5.5.2 Influence of the modification by formaldehyde

The influence of the strength of the formaldehyde treatment (compare

Chapter 6) was investigated using the conditions outside the transfer

limitation region for hydrogen. The selectivity towards butylamine was

higher if a low concentration of formaldehyde was used to modify the

catalyst (from 88.6 mass-% to 93.3 mass-%, compare Figure 5-25). The

hydrogénation rates decrease dramatically, so that if higher formaldehyde

concentrations are used to modify the catalyst, the hydrogénation reactions

Page 63

100-1

butylamine

dibutylamine

Bli tributylamidine

I 1 max. dibutylimine

unmodified 1% 2% 3.5%

[ChLO] / [mass-%]

Fig. 5-25 : Influence of the concentration of formaldehyde used to modify Raney nickel

on the product distribution. Reaction conditions: 240 g BN, 3.75 g RaNi,

100°C, 1 MPa and 1000 rpm.

are so slow, that the dimerisation can compete successfully the

hydrogénation (Table 5-3 and Figure 5-26). In Figure 5-26 the relative

initial reaction rates that were calculated by dividing the initial rates

obtained with the modified catalyst by the initial rates of the unmodified

catalyst are compared for various modification procedures. The reaction

leading to DBI (d[DBI]/dt) is somewhat stronger decelerated than the

reactions leading to butylamine.

Figure 5-27 reveals that the measured data points for the modified

catalysts do not fit the correlation between the yield butylamine versus the

initial differential selectivity for the unmodified catalysts shown in

Figure 5-5. Thus, the improved selectivity behavior of the modified catalyst

can not be solely explained by mass and heat transport limitation effects.

If modified nickel-on-carrier is used at the same conditions, again a rise

in selectivity is observed (Figure 5-28). The benefit of the modification

(from 84.1 mass-% to 92.1 mass-%) is even more impressive as in the case

of Raney nickel. Decreased hydrogénation rates were observed again, if the

Page 64

catalyst was modified (Table 5-4 and Figure 5-29). Plotting the relative ini¬

tial reaction rates again shows a stronger deceleration of the reaction leading

to dibutylimine.

Table 5-3: Influence of the modification strength on the initial reaction rates. Reaction

conditions: 240 g BN, 3.75 g RaNi, 100°C, 1 MPa and 1000 rpm.

modification

unmodified

1% CH20

2% CH20

3.5%CH20

5% CH20

-d[BN]/dt d[BA]/dt d[DBA]/dt d[DBI]/dt

/ [mmol/s*kg] / [mmol/s*kg] / [mmol/s*kg] / [mmol/s*kg]

118.656

82.262

42.310

18.743

12.233

82.612

59.926

31.967

13.771

8.806

3.199

2.322

1.034

0.455

0.314

14.258

8.194

3.633

1.835

1.276

-d[BN]/dt / [118.656 mmol/(s*kg)]

d[BA]/dt / [82.612 mmol/(s*kg)]

d[DBA]/dt/ [3.199 mmol/(s*kg)]

d[DBI]/dt / [14.258 mmol/(s*kg)]

unmodified 1% 2% 3.5%

[CH20] / [mass-%]

5%

Fig. 5-26: Influence of the formaldehyde modification of Raney nickel on the relative

initial reaction rates. Reaction conditions: 240 g BN, 3.75 g RaNi, 100°C,

1 MPa and 1000 rpm.

Page 65

94-

92-

toau¬

to03

"88-

CD ^ various reaction

CDCO

!LUB|Aj parameters (see Fig 6-3)

.q 84- • 1%CH202% CH203.5% CH205% CH20

o

CD

's* 82-

80-

1 '

10

1 '

20

1

30

i ' i

40 50

initial differential selectivity / [-]

Fig. 5-27: Initial differential selectivity versus product distribution (yield butylamine).

100 -,

95-

<n

E

o

'l_

W

T3

T3

O

E3

butylamine

dibutylamine

tributylamidine

max. dibutylimine

unmodified 2.5%

[CKO] / [mass-%]

Fig. 5-28: Influence of the formaldehyde modification of nickel-on-carrier on the

product distribution. Reaction conditions: 240 g BN, 10 g nickel-on-carrier,

100°C, 1 MPa and 1000 rpm.

Page 66

Table 5-4: Influence of the formaldehyde modification of nickel-on-carrier on the initial

reaction rates. Reaction conditions: 240 g BN, 10 g nickel-on-carrier, 100°C,

1 MPa and 1000 rpm.

modification -d[BN]/dt d[BA]/dt d[DBA]/dt d[DBI]/dt

/ [mmol/s*kg] / [mmol/s* kg] / [mmol/s*kg] / [mmol/s*kg]

unmodified 30.605 15.979 1.430 6.219

2.5% CH20 14.867 10.673 0.448 1.542

5% CH20 8.810 5.734 0.256 1.256

unmodified

-d[BN]/dt/ [30.605 mmol/(s*kg)]

d[BA]/dt / [15.979 mmol/(s*kg)]

d[DBA]/dt / [1.430 mmol/(s*kg)]

d[DBI]/dt/ [6.219 mmol/(s*kg)]

2.5%

[CH O] / [mass-%]

Fig. 5-29: Influence of the formaldehyde modification of nickel-on-carrier on the

relative initial reaction rates. Reaction conditions: 240 g BN, 10 g nickel-on-

carrier, 100°C, 1 MPa and 1000 rpm.

Page 67

5.6 Reversibility of the hydrogénation steps

5.6.1 Disproportionation of butylamine

To investigate the reversibility of the hydrogénation steps, experiments

using butylamine and dibutylimine as starting materials were carried out. In

a first experiment, the conditions were identical with those used as standard

conditions to hydrogenate butyronitrile (254 g BA, 24 g RaNi, 100°C,

1 MPa, and 1000 rpm). The reaction profile is shown in Figure 5-30. The

100

90

80

^ 70

COCO 60cc

F50

c

o 40

CO

oQ.

30

EoÜ

20

10

butylamine

dibutylamine

tributylamine

~i

10

time / [day]

Fig. 5-30: Reaction profile using butylamine as starting material. Reaction conditions:

254 g BA, 24 g RaNi, 100 °C, 1 MPa and 1000 rpm.

first derivation of the concentration profile results in the disproportionation

rates of the primary amine at a given concentration. Using this experiment,

the reversibility of the hydrogénation reaction leading from the primary

amine to the secondary amine is proved. This experiment also shows that

the amount of the primary amine obtained by the catalytic hydrogénation of

Page 68

butyronitrile with Raney nickel is mainly determined by the kinetics and not

by the thermodynamics (equilibrium) of the process. This fact may also

explain the obtained low selectivity of the third cycle when the recycling of

the catalyst was investigated (Figure 5-16) or the low selectivity of the

catalyst modified by 5% formaldehyde. If the hydrogénation is to slow, i.e.

if the reaction time is to high, the obtained selectivity towards butylamine is

not the maximum possible one (kinetic) but strives towards the much lower

thermodynamically determined product distribution. This is shown

in Figure 5-31.

100

80-

co

CO

| 60

w 40oQ.

EoÜ

20

—— butylamine

• butyronitrile* dibutylimine

——dibutylamine

WW*---

>- » » I—•" 1 //—100 200 300 1000

time / [min]

1100 1200

Fig. 5-31: Hydrogénation of butyronitrile (not stopped after full conversion). Reaction

conditions: 240 g BN, 3.75 g RaNi, 100 °C, 1 MPa and 1000 rpm.

A second experiment was performed, using dibutylamine as starting

material. 240g DBA, 20 g RaNi and 30 g NH3 were placed in the reactor.

The stirrer speed was set at 1000 rpm and the temperature was 100°C. No

reaction was observed, after 5000 min. The explanation for this behavior is,

that the dibutylamine could not make sufficient contact with the catalytic

surface of the wet catalyst.

Page 69

5.6.2 Influence of reaction parameters on the disproportionation of

butylamine

The influence ofthe temperature and the hydrogen pressure was investigated

using the 35 ml reactors (see Chapter 9.1.4) and the procedure described in

Chapter 9.2.2. The reactors were charged with high catalyst loadings, and

the hydrogen was pressed into the reactor before heating. The pressure

during the reaction was not recorded. It is evident, that the temperature has a

great influence on the disproportionation rate (Figure 5-32).

— 100 °C

• 120°C

* 140°C

——160°C

U-| 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

time / [day]

Fig. 5-32 : Influence ofthe temperature on the disproportionation rate ofbutylamine. Re¬

action conditions: 20 g BA, 2 g RaNi, 280 min."1 without hydrogen.

The hydrogen pressure also influences the disproportionation rate.

Higher hydrogen pressures decrease the reaction to the secondary amine

(Figure 5-33). A dehydrogenation of butylamine as the first step in the

disproportionation mechanism could be an explanation for this behavior [41].

As mentioned above, a faster disproportionation of butylamine can

decrease the high yield obtained under kinetic controlled conditions. This

Page 70

110-1

wCO

CD

E

cu

g

ETO

.Q

without H„

2 MPa H24 MPa H„

5 6

time /[day]

11

Fig. 5-33: Influence of the hydrogen pressure on the disproportionation of butylamine.

Reaction conditions: 20 g BA, 2 g RaNi, 120°C and 280 min."1.

CO

CO

CD

E

110-,

100-

90-

80-

70-

60-

cu

ETO

X2

50-

40-

30-

20-

10-

-0 1

unmodified

1%CH20 modified

2% CH20 modified

3.5% CH20 modified)

5% CH20 modified

00

—r~

01

—r~

02 03

time / [day]

—i—

04 05

—I—

06

—I

07

Fig. 5-34: Influence of the catalyst modification on the disproportionation of butyl¬

amine. Reaction conditions: 20 g BA, 2 g RaNi, 120°C and 280 min."1.

Page 71

argumentation was also used to explain the lower selectivity, if a catalyst,

modified by 5% formaldehyde was used. The influence of the catalyst

modification by formaldehyde on the disproportionation rate is shown

Figure 5-34.

5.6.3 Dibutylimine as starting material

Dibutylimine was synthesised from butylamine and butyraldehyde using an

excess of butylamine. 174 g of the obtained solution, containing 69.5%

dibutylimine (0.95 mol), 12.4% butylamine and 18.1% side products from

the synthesis was used for the disproportionation study. The reaction

conditions were: 3.79 g RaNi, 20 g NH3 (1.18 mol), 100°C, 2 MPa,

1000 rpm. During the heating period dibutylimine already disappears and

butyronitrile and butylamine are produced This experiment shows that the

hydrogénation of butyronitrile is reversible too.. A second result is, that the

COCO

CD

E

o

"55oQ.

Eoo

90-1

80-

70-

60-

50-

40-

30-

20-

10-

0-

-— butylamine-•— butyronitrile*— dibutylimine-— dibutylamine

side products

^ y w w ^ w'—iip • |p ~— up ~— up —-

-50

T

50 100 150 200 250 300 350 400

time / [min]

Fig. 5-35: Reaction profile using dibutylimine as starting material. Reaction conditions:

174 g DBI, 3.75 g RaNi, 20 g ammonia, 100°C, 2 MPa and 1000 rpm

Page 72

amount of dibutylimine is not decisively influencing the selectivity of this

reaction, or in other words: the reaction rates transforming dibutylimine,

butylamine and butylimine to the state of equilibrium are faster than the

reaction rate of the hydrogénation of butylimine leading to butylamine

(Figure 5-35).

5.7 Discussion

5.7.1 The bifunctional catalytic hydrogénation and its reversibility

The mechanism presented in Chapter 3.1 (Figure 3-1) postulates, that the

selectivity of the hydrogénation of nitriles depends on two catalytic

functions, an acidic and a hydrogénation function. The ratio and the distance

between these two functions is assumed to determine the ratio of the

reaction constants kH4/kC5 (in Figure 3-1) and therefore the yield of primary

amine. Furthermore, the maximum produced amount of dibutylimine was

observed to be higher than the produced amount of dibutylamine at full

conversion, so that a reaction path from dibutylimine to butylamine is

highly probable (see Figure 5-9, Figure 5-13 and Figure 5-31). From the

experiments using dibutylimine and ammonia as starting material it is

obvious, that high selectivities towards the primary amine can be obtained,

even if the condensation products are already present in the reaction

mixture. This leads to the conclusion, that dibutylimine is in a fast

equilibrium with ammonia and a species that can be hydrogenated to the

primary amine. In such disproportionation experiments with the butylamine

or the condensation by-products as starting materials butyronitrile as

intermediate is detected. This reveals the reversibility of the hydrogénation

as well as the condensation steps. Depending on the reaction conditions

applied, such reversible reaction systems can exhibit, as two extremes, a

kinetically or a thermodynamically determined product distribution. The

catalytic hydrogénation of the aliphatic butyronitrile is obviously such a

reaction system with butylamine as the kinetically controlled product. Thus,

Page 73

in order to optimize the selectivity towards butylamine, the reaction must be

well monitored over time and immediately stopped after conversion has

been completed. If the reaction proceeds beyond this time the butylamine

disproportionates and the reaction strives towards a product distribution

determined by the equilibrium. This, however, is unfavourable for a

selective production of butylamine.

5.7.2 Influence of various reaction parameters on the selectivity

First attempts to get a qualitative understanding of the influence of various

reaction parameters on the selectivity towards the desired primary amine

failed due to the complexity of the hydrogénation system. Therefore, a

simple semi-quantitative macro-kinetic model has been derived, with the

help of which the complex selectivity behavior of the explored reaction

system could more easily be characterised. For the sake of simplicity the

rate constants of the adsorption and desorption steps of the various reaction

components to and from the catalyst surface, respectively, has in each case

been incorporated into the overall rate constant of the corresponding

reaction step. The explanation of the effects of various reaction parameters

such as the gas-liquid transfer limitation of hydrogen, the overall pressure,

the temperature, the amount and recycling of catalyst and the presence of

various additives on the yield of the desired butylamine led to the

conclusion that all measured parameter effects can be explained if the

following question can be answered: "What influence do the changes of the

various reaction parameters exert on the local gradients of temperature and

hydrogen concentration near the catalyst surface?" It has been shown

(Chapter 5.4) that the selectivity towards butylamine increases if the local

temperature increases (Chapter 5.4.2) or if the local hydrogen concentration

decreases (Chapter 5.4.1). This can be rationalised with the help of the

"quasi" Arrhenius plot (Figure 5-12) and the equation for the differential

selectivity (Eq. 5.10) taking also into account aspects of mass and heat

transfer effects (Chapter 5.3). The more these local gradients change by

changing the reaction parameters the lower the influence on the selectivity

Page 74

becomes. Therefore, the most favourable selectivity towards butylamine can

be expected, if highly active catalysts are used and as a consequence, a

temperature and mass diffusion limitation becomes dominant causing high

local gradients.

Page 75

Page 76

Chapter Vf

Modification of nickel catalysts by

formaldehyde

6.1 General remarks

The modification mechanism of nickel or cobalt catalysts by formaldehyde

is still unknown. The interaction of formaldehyde and/or products of its

decomposition or polymerisation with the catalyst, produces a catalyst with

different properties, thereby producing higher selectivities towards the

primary amines, if nitriles are hydrogenated [1, 63, 64, 81]. One explanation

for the interaction of formaldehyde and the nickel catalyst is a

disproportionation of formaldehyde to methanol and carbonmonoxide

(Eq. 6.1).

2 CH20 CH3OH + CO Eq. 6.1

The formed methanol desorbs from the catalyst while carbonmonoxide is

still adsorbed on the catalysts surface as Nix(CO)y. This mechanism would

be in agreement with reported results, where the interaction of

formaldehyde with Ni(llO) surfaces was investigated at low temperatures

(95 K) by Richter and Ho [82]. Formaldehyde reacts on the catalysts

surface, producing a mixed CH30 and CO adlayer on the surface (Eq. 6.2).

CH20 + CH20 CH30(a) + CO(a) + H(a) Eq. 6.2

The adsorbed species that arised due to the adsorption and thermal

processing of CH20 on Ni(llO) were summarised in Eq. 6.3-Eq. 6.7. Mixed

Page 77

paraformaldehyde and solid H2CO multilayers were formed at high

formaldehyde exposures.

CH20(g) CO(a) + 2 H(a) Eq. 6.3

2 CH20(g) CH30(a) + CO(a) + H(a) Eq. 6.4

n CH20(g) (CH20)n(a) Eq. 6.5

(CH20)n(a) HCOO(a) + CH30(a) + CO(a) + H(a) Eq. 6.6

+ CHx(a) + CH20(g)

CH30(a) *- CO(a) + 3 H(a) Eq. 6.7

If the catalyst was heated after the formaldehyde treatment, methanol

desorbed from the catalyst at temperatures of about -10°C, carbonmonoxide

desorbed at temperatures above 174°C.

Newton and Dodge [83] investigated the equilibrium constants between

carbon monoxide, hydrogen, formaldehyde and methanol (Eq. 6.8 and

Eq. 6.9).

CO + H2 ^=^ CH20 Eq. 6.8

CH20 + H2 =*=^ CH3OH Eq. 6.9

The scope of their work was to investigate, if it is possible to produce

formaldehyde by hydrogénation of carbon monoxide. One of the results

was, that nickel catalysts promote the decomposition of formaldehyde into

carbon monoxide and hydrogen. At 200°C and 0.1 MPa the equilibrium

constants were Kj = 2.30 * 10"5 (Eq. 6.8) and K2 = 1800 (Eq. 6.9). The

conclusion was, that the production of formaldehyde from carbon monoxide

by hydrogénation is not feasible at any reasonable temperature or pressure.

Several investigations concerning the chemisorption of CO on different

Ni surfaces were recently made [84-87]. Especially the adsorption mode

Page 78

was investigated, thereby two species were observed: Bridged CO and

on-top or terminal CO. Both species coexist and are in an equilibrium

depending on pressure (surface coverage) and temperature. The desorption

temperature is in the range of 170°C.

The washing operations as well as the formaldehyde treatment were

made in a three-neck sulfonation flask (see Chapter 9.1.5) as described in

Chapter 9.2.7-9.2.10.

Modification experiments were usually made, using the procedures of

Degischer and Rössler [1, 63]. A 5% formaldehyde modification of a Raney

catalyst means, that X g Raney nickel are modified by 2X g of an aqueous

solution containing 5% formaldehyde (normal procedure: 50 g Raney nickel

in 100 g aqueous solution, that contains 5 g formaldehyde).

Methanol (by headspace-GC), formaldehyde (by HPLC) and the nickel

concentration (by XRF) were determined in the aqueous modification

solution after a modification time of 30 min.

6.2 Influence of the treatment with different solvents on the

properties of the catalyst

6.2.1 Reduction potential

The standard reduction potential of fresh, water washed, alcohol and

formaldehyde modified catalysts was measured as described in

Chapter 9.5.1. The potential of the fresh and the water washed catalyst is in

a range of -0.6 V indicating that the water washing does not change the

properties of the catalyst. This observation is in agreement with the results

of the nitrile hydrogénation (see Figure 5-19 and Figure 5-20). The

potential of the methanol, ethanol and formaldehyde modified catalyst is in

a range of -0.3 V, indicating that these substances do change the catalysts

properties. In addition, if the fresh catalyst was washed with water or

methanol, no nickel was found in the washing solutions.

Page 79

650-

600-

550-

^ 500-

£ 450-

öj 400-

S 350-

Q. 300-

1 250-

T3c 200-

3« 150-

I

100-

50-

0- -1 '—'—' 1 '—'—' 1 '—'—' 1-


Fig. 6-1: Standard reduction potential of fresh and modified catalysts determined with

a combined gold electrode.

6.2.2 Adsorption of an indicator

In order to get some information about the influence of the modification

treatments on the acid sites activity the adsorption capacity for

4-aminoazobenzene was determined for fresh, water washed, methanol and

formaldehyde modified catalysts as described in Chapter 9.5.2. Four

measurements were carried out. Figure 6-2 reveals that the acid activity of

the catalysts does depend on the modification treatment applied. However,

no simple correlation between these measurements and the selectivity data

was found.

Page 80

40-

35-

, ,

n>

o 30-

F^r

"

O2 b-

'—'

—..

>s

75 20-

roQ.

mo 1 5-

c

o-!—«

Q. 1 0-

O<n

"

n

ro 05-

00-

fresh 3 x H20 3 x MeOH 5% formaldehyde

Fig. 6-2 : Adsorption of 4-aminoazobenzene on Raney nickel.

6.3 Modification of Raney nickel by various formaldehydeconcentrations

6.3.1 Analysis of the modifying solution

Different formaldehyde concentrations were used (1%, 2%, 3.5%, 5% and

7% in water), while the mass of catalyst and the mass of the aqueous

solution were not varied (50 g wet Raney nickel and 100 g solution). The

methanol, formaldehyde (which was not converted during the modification)

and residual formaldehyde amounts are plotted in Figure 6-3. At low

formaldehyde concentrations, formaldehyde disappears completely, at

higher concentrations not all formaldehyde is consumed. A remarkable

result is, that the amount of methanol corresponds to the half amount of the

consumed formaldehyde at low modification concentration (Table 6-1). In

this table, the mass balance of the modification procedure is given, as well

as the division of methanol found by the residual mass in [mmol CH20] that

was not found. At low concentrations, the stoichiometry of methanol and

Page 81

[CH.O] / [mass-%]

Fig. 6-3: Amount of methanol, formaldehyde and residual mass in the modifying

solution at various modification conditions. Reaction conditions: 50 g Raney

nickel modified in total 100 g aqueous solution.

Table 6-1: Mass balance of the catalyst modification by various formaldehyde

concentrations. Initial formaldehyde, not converted formaldehyde and

methanol determined in the modification solution and the residual mass

(volume: 100 ml).

initial CH20 MeOH residual MeOH/ hydrogen needed

CH20 measured measured residual to produce MeOH

/ [mmol] / [mmol] / [mmol] / [mmol CH20] /[-] / [ml/g cat]

33.9 0.0 17.0 17.0 1.00 7.6

68.0 0.0 29.4 38.6 0.76 13.2

118.1 5.9 51.6 60.7 0.85 23.1

169.2 11.7 87.3 70.2 1.24 39.5

237.0 66.4 87.6 83.0 1.06 39.2

Page 82

residual mass is 1, indicating that formaldehyde reacts according Eq. 6.1 to

methanol and carbon monoxide. The last row of the table, describes the

hydrogen, that would be needed to produce the measured methanol from

formaldehyde by direct hydrogénation. This value was 39 ml hydrogen per

g catalyst at high modification concentrations. Literature values for the

hydrogen content stored on Raney nickel are about 20 ml/g, so that not

enough hydrogen is stored on the catalyst to produce the methanol measured

in the modification solution [45].

The nickel concentration in the liquid phase after modification is shown

in Figure 6-4. The higher the formaldehyde concentration is, the higher is

the nickel amount removed from the catalyst. There exists almost a linear

correlation between the modification concentration and the leached nickel.

5000-

4000-

EQ.

B: 3000•

o

^ 2000-

CD.co

CD

_ÇD

1000 4

[CH20] / [mass-%]

Fig. 6-4: Concentration of leached nickel in the modifying solution determined by

XRF. Reaction conditions: 50 g Raney nickel in total 50 g aqueous solution.

Page 83

6.3.2 Properties of the modified catalysts

The standard reduction potential of the modified catalysts are presented in

Figure 6-5. The potential is higher, if the catalyst is modified by

600 -,

500-

•>-

E400-

-~~-

CD .

.|_«

c

CD 300-oQ.

T3i_

coT3

200-

C

m

w

100-

0 —'—I—! H-1 H-1 H 1—I—I H H 1 *—I—*

01 2345678

[CH20] / [mass-%]

Fig. 6-5: Standard reduction potential of Raney nickel modified by various

formaldehyde concentrations determined with a combined gold electrode.

formaldehyde. A value of -0.35 V is obtained for catalyst, treated by high

formaldehyde concentrations.

The stability against acidic attack is shown in Figure 6-6. The fresh

catalyst is completely dissolved in hydrochloric solution, while the 7.5%

formaldehyde modified catalyst at these conditions is stable and almost

releases no hydrogen.

Page 84

350-,

1

300-

(/)

>s -

m

m 250-o

n) -

E 200-

—.

T3"

CD

ce150-

fl)

CDi_

c 100-CDO)ol_

T3 50-

.c

°H—>—i—'—i—'—i—'—i—'—i—'—i—'—i—'—i01 2345678

[CH20] / [mass-%]

Fig. 6-6: Hydrogen evolved by treating Raney nickel catalysts with hydrochloric acid.

Reaction conditions: 1 g Raney nickel was dissolved in 70 g of a 10%

hydrochloric acid.

6.4 Modification of various amounts of Raney nickel at

constant modification strength

6.4.1 Analysis of the modifying solution

Different catalyst amounts (12.5 g, 25 g, 37.5 g, 50 g, 75 g and 100 g) were

used while the formaldehyde concentration and the mass of the aqueous

solution were constant (113 g of a 4.5% formaldehyde solution). The

methanol and residual formaldehyde amounts are plotted in Figure 6-7. The

amount of disappeared formaldehyde, that is not found as methanol in the

liquid phase is plotted as residual mass. At low catalyst amounts, not all the

formaldehyde reacts. At higher catalyst amounts, all formaldehyde

disappears. The same calculation as in Chapter 6.3 was made again and the

results are listed in Table 6-2. At high catalyst masses (low formaldehyde

concentration / catalyst ratio), the stoichiometry between methanol found in

Page 85

40 60 80 100

amount catalyst in a 4.5% formaldehyde solution / [g]

Fig. 6-7: Amount of methanol, formaldehyde and residual mass found in the modifying

solution. Reaction conditions: 113 g of an aqueous solution containing 4.5%

formaldehyde.

Table 6-2: Mass balance of the catalyst modification by various catalyst amounts. Mass

of catalyst used, not converted formaldehyde and methanol determined in the

modification solution and the residual mass (initial formaldehyde amount

169.2 mmol, volume: 113 ml).

catalyst CH20 MeOH residual MeOH/ hydrogen needed

amount measured measured residual to produce MeOH

/[g] / [mmol] / [mmol] / [mmol CH20] /[-] / [ml/g cat]

12.5 119.2 14.2 35.9 0.39 25.1

25 92.9 30.8 45.4 0.68 27.2

37.5 47.7 55.1 66.5 0.83 33.0

50 49.6 57.2 62.6 0.91 25.5

75 0 86.5 82.7 1.05 25.84

100 0 86.9 82.5 1.05 19.5

Page 86

the solution and residual mass is almost 1, at low amounts of catalyst, the

stoichiometry is far from 1.

The nickel concentration in the liquid phase after modification is shown

in Figure 6-8. The more catalyst being present in the modifying system, the

more nickel is leached from the catalyst.

3000-

2500-

E 2000-Q.Q.

CD

ü

oCD

Ücc

CD

1500-

1000-

500-

20 40 60 80 100


Fig. 6-8: Concentration of leached nickel in the modifying solution determined by

XRF. Reaction conditions: 113 g of an aqueous solution containing 4.5%

formaldehyde modifying various amounts of catalyst.

6.4.2 Properties of the modified catalysts

The standard reduction potential of the modified catalysts is plotted in

Figure 6-9. The amount of hydrogen which is released if the catalyst is

attacked by hydrochloric acid is shown in Figure 6-10. If the formaldehyde

modification is carried out with small amounts of catalyst the released

hydrogen remains the same.

Page 87

400-

350-

300-

rö 250 4

CD

"5 200-Q.

T3

ro lOO-

I 100.

50-

+- -+-

20 40 60 80 100


Fig. 6-9: Standard reduction potential of various catalyst amounts modified by a 4.5%

formaldehyde solution.

180-

160-

CO>< 140cch—»

cco 120O)

E100

^

c

CD

O)80

O&_

T>

>< HO.C

oCDCO 40

CCCD

P 20-

2080

40 6080

100


Fig. 6-10: Hydrogen evolved from differently modified catalysts after treatment with

hydrochloric acid. Reaction conditions: 1 g Raney nickel dissolved in 70 g of

a 10%) hydrochloric acid.

Page 88

6.5 Modification of nickel-on-carrier

The modification procedure was not only investigated using Raney

catalysts, but also using a nickel-on-carrier as hydrogénation catalyst. If the

catalyst was modified by a 5% formaldehyde solution (169 mmol

formaldehyde), not all formaldehyde was consumed (Table 6-3). If

unmodified catalyst is washed with water, no nickel is leached from the

catalyst (< 10 ppm). However, if the catalyst was modified by

formaldehyde, a concentration of 463 ppm nickel was found in the

modification solution.

Table 6-3 : Mass balance of the modification of nickel-on-carrier. Mass of catalyst used,

formaldehyde and methanol determined in the modification solution and the

residual mass.

catalyst

amount

/[g]

CH20

measured

/ [mmol]

MeOH

measured

/ [mmol]

residual

/ [mmol CH20]

MeOH/

residual

/[-]

hydrogen needed

to produce MeOH

/ [ml / g cat]

50 113.3 19.8 36.0 0.55 8.8

6.6 Discussion

The first conclusion which can be drawn from modification experiments

with formaldehyde is that changes of the catalyst properties due to water

washing operations before and after the modification by formaldehyde can

be excluded.

Up to 50% of the formaldehyde used to modify the catalyst can be

found as methanol in the solution. If modification conditions were used

whereby more than 50 g catalyst amount and a formaldehyde concentration

of 1 mass-% were applied almost exactly 50% were found as methanol. This

is an indication that the reaction of formaldehyde to methanol together with

Page 89

an adsorbed species occurs. The catalyst does not store enough hydrogen to

hydrogenate formaldehyde to the produced amount methanol. Therefore,

this additional hydrogen has to be taken from somewhere else presumedly

from formaldehyde itself or from the solvent. These facts are strong

indications that formaldehyde reacts according to Eq. 6.1 at low

concentrations of formaldehyde and high amounts of catalyst. The

production of chemisorbed carbon monoxide would also explain the effect

of other modifiers, such as acetaldehyde, benzaldehyde, carbon dioxide,

acetone and carbon monoxide itself that were described in the patent [1].

From all these modifiers, carbon monoxide could be produced on the

surface of the catalyst.

The effect of chemisorbed carbon monoxide on the hydrogénation of

nitriles could be explained by stereochemical restrictions. Carbon monoxide

blocks metal sites on the catalyst, so that larger substrates (dialkylimines)

can not adsorb as easily as smaller substrates (alkylimines). Tributylamidine

was observed as by-product if modified catalysts were used to hydrogenate

butyronitrile. One explanation for the formation of this product, is that this

large molecule can not be hydrogenated to tributylamine at the chosen

conditions because of stereochemical reasons.

Large amounts of nickel can be found in the modification solution, an

important fact, if modifications have to be made on a larger scale. The

minimization of this nickel leaching during the modification procedure and

also during the hydrogénation reaction is evidently important for the

preparation of large amounts of catalyst.

The catalysts treated by high formaldehyde concentrations are almost

stable in hydrochloric acid, while a fresh catalyst is completely dissolved in

a short time.

Page 90

Chapter /

Effect of formaldehyde modified nickel

catalysts on other chemical systems

Formaldehyde modified nickel catalysts have revealed an unexpected and

positive selectivity effect on the hydrogénation of nitriles (Chapter 5.5). The

exact reason for this observation is not yet known and its investigation shall

be the aim of future research studies. In order to examine whether these new

catalysts also bring about such selectivity effects in other well known

hydrogénation processes, in this Chapter some typical hydrogénations were

screened.

7.1 Hydrogénation of crotonaldehyde

7.1.1 General remarks

The hydrogénation of oc,ß-unsaturated aldehydes is still a challenging field

of investigation. The desired oc,ß-unsaturated alcohol is not the

thermodynamic product, and therefore, the saturated aldehyde is preferably

formed. A change of the selectivity has to be achieved by a change of the

reaction rate constants of the competitive reactions and of the competitive

adsorption constants of the components. The catalytic system that produces

the highest selectivities are Pt/Ti02, Pt/Si02, doped by different transition

metals or modifiers [88-91]. Selectivities up to 50% were achieved if

crotonaldehyde was hydrogenated in ethanol. NiPt/Si02 catalysts were also

tested, but produced only small amounts of the desired alcohol [92].

Page 91

Selectivities up to 85% towards the unsaturated alcohol can be achieved

using bimetallic Ag/Si02 catalysts [93, 94] or Ru/Si02 catalysts [95, 96].

Investigations on the influence of the catalyst modification on the

hydrogénation of oc,ß-unsaturated aldehydes were performed in a system of

crotonaldehyde and ethanol as solvent. A scheme of the reaction system is

given in Figure 7-1. E/Z crotonaldehyde reacts to butanol over butanal or

over E/Z crotylalkohol as intermediate.

E/Z crotylalkohol

Fig. 7-1 : Hydrogénation of E/Z crotonaldehyde to butanol over the intermediates

butanal or E/Z crotylalkohol.

Experiments were made in a 500 ml steel hydrogenator (see

Chapter 9.1.1) according the procedure described in Chapter 9.2.3, and

samples were taken according the procedure in Chapter 9.2.6.

1,1-diethoxybutane was observed as by-product (Figure 7-2) or even as

main product, in cases that Raney nickel was modified by formaldehyde.

This product is formed in an acid catalysed side reaction via the semiacetal

by addition of the solvent and subsequent dehydration [90].

Page 92

.0. 1,1 -diethoxybutane

CL /

Fig. 7-2: 1,1-diethoxybutane was formed as by-product, if crotonaldehyde was

hydrogenated.

7.1.2 Test for a possible gas-liquid transfer limitation for hydrogen

To investigate whether a gas-liquid transfer limitation for hydrogen exists at

the chosen reaction conditions, the amount of catalyst was doubled (40 g

E/Z crotonaldehyde, 200 g ethanol, 30°C, 1 MPa overall pressure and

1000 rpm). As can be seen in Table 7-1 and Figure 7-3 the reaction rate

increases as the catalyst amount was raised, so that a transfer limitation can

be excluded.,

Table 7-1 : Influence of the catalyst amount on the initial reaction rates, if E/Z croton¬

aldehyde is hydrogenated. Reaction conditions: 40 g E/Z crotonaldehyde,

200 g ethanol, 30°C, 1 MPa and 1000 rpm.

catalyst initia disappearance rate of initial production rate of

amount crotonaldehyde butanol

/[g] / [mmol/s] / [mmol/s]

2 0.193 0.0024

4 0.395 0.0049

Page 93

100-,

COcoce

COoQ.

OÜ

75

time / [min]

150

Fig. 7-3: Influence of the catalyst amount on the hydrogénation rates. Reaction

conditions: 40 g crotonaldehyde, 200 g ethanol, 30°C, 1 MPa and 1000 rpm.

7.1.3 Influence of the formaldehyde modification ofRaney nickel on the

hydrogénation of crotonaldehyde

Modification of Raney nickel with formaldehyde leads to a lower

disappearance rate of crotonaldehyde and a lower production rate for

butanol (Figure 7-4 and Figure 7-5). Butanal is the only observed

intermediate, no crotylalkohol was observed during the reaction. If the

catalyst is modified using a 5% formaldehyde solution, condensations take

place and 1,1-diethoxybutane is observed as main product.

Page 94

COCOce

oü

100

80-

60-

unmodified

1 % CH202% CH203.5% CH20

S 40-cooQ.

20-

100 125 150 175 200 225 250 275 300

time / [min]

Fig. 7-4: Influence of the modification strength of the catalyst on the crotonaldehyde

disappearance rate. Reaction conditions: 40 g crotonaldehyde, 2 g RaNi,

30°C, 1 MPa and 1000 rpm.

20-1

18-

16-

-9-

o^ 14-

CO

CO

CO 12-

E -

10-

c

o

-I-»8-

CO .

o

Q. 6-

E -

o 4-o

—— unmodified, butanol

—•—-1%CH20 modified, butanol

—A—- 2% CH20 modified, butanol

——- 3.5% CH20 modified, 1,1-diethoxybutane- 3.5% CH20 modified, butanol

—+—- 5% CH20 modified, 1,1-diethoxybutaneX—- 5% CH20 modified, butanol

1 r^T

150 200

T—' 1—' 1 '—I—' 1—' 1 '—I—' 1

250 300 350 400 450 500 550 600

time / [min]

Fig. 7-5: Influence of the modification strength of the catalyst on the product

distribution and formation rate. Reaction conditions: 40 g crotonaldehyde, 2 g

RaNi, 30°C, 1 MPa and 1000 rpm.

Page 95

7.1.4 Influence of the formaldehyde modification of nickel-on-carrier

on the hydrogénation of crotonaldehyde

The influence of the modification by formaldehyde on the selectivity and

activity was also investigated using a nickel-on-carrier catalyst. The

behavior of the system was as in the case of Raney nickel (Figure 7-6). The

0 50 100 150 200 250 300 350 400

time / [min]

Fig. 7-6: Influence of the modification of a nickel-on-carrier catalyst on the reaction

rates. Reaction conditions: 40 g crotonaldehyde, 4 g nickel-on-carrier, 30°C,

1 MPa and 1000 rpm.

modification lowers the hydrogénation rate of crotonaldehyde if

formaldehyde concentrations above ca. 2.5% for the modification were

used. No 1,1-diethoxybutane was observed as by-product using this catalyst.

Page 96

7.2 Hydrogénation of l-bromo-4-nitrobenzene

7.2.1 General aspects

Several possibilities exist to reduce nitroaromatic compounds. The

Béchamps-reduction, the reduction using other metals than iron, the

reduction by sulphides, electrochemical reductions or the reduction using

hydrazine [97].

The hydrogénation of nitroarenes, especially of halide substituted

arènes using Raney nickel as catalyst is of industrial interest [98]. One of the

problems thereby is the dehalogenation of the substrate or product.

Therefore, there was some hope that the newly discovered formaldehyde

nitrobenzene

Fig. 7-7: Reaction scheme for the hydrogénation of l-bromo-4-nitrobenzene to

l-bromo-4-aminobenzene and the undesired side-reactions (hydrogenolysis

of bromine) to aniline.

Page 97

modified catalyst might prevent the undesired dehalogenation. The

hydrogénation of a halide substituted nitroarene was investigated using the

substrate l-bromo-4-nitrobenzene. A reaction scheme is given in

Figure 7-7. The experiments were made in a 200 ml glass hydrogenator

(Chapter 9.1.2) according the procedure described in Chapter 9.2.4. After a

modification procedure the catalyst was washed three times with

tetrahydrofurane (Chapter 9.2.9). The reaction conditions were: 0.5 g Raney

nickel, 10 g l-bromo-4-nitrobenzene, 90 g THF, 50°C, 0.5 MPa and 1500

rpm.


To check whether an influence of the gas-liquid transfer limitation for

hydrogen exists, the hydrogénation rate was doubled by using twice the

amount of the catalyst (Table 7-2). This is an indication that no limitation

Table 7-2: Selectivity and rate of the hydrogénation of l-bromo-4-nitrobenzene.

Reaction conditions: 10 g l-bromo-4-nitrobenzene, 90 g THF, 50°C, 0.5 MPa

and 1500 rpm.

Raneynickel initial hydrogen initial l-bromo-4- aniline

amount uptake hydrogénation rate aminobenzene

/[g] / [mmol/s] / [mmol/(s*kg)] / [mass-%] / [mass-%]

0.48 0.011 22.2 99.3 0.7

1.00 0.028 28.0 98.5 1.5

for hydrogen exists. In addition, if the catalyst was not washed with

tetrahydrofurane, no reproducible reaction rates were obtained.

Page 98

7.2.3 Influence of the modification on selectivity and reaction rates

The selectivity towards l-bromo-4-aminobenzene is lower if the catalyst is

modified by formaldehyde and higher again, at high modification

concentrations of formaldehyde (Figure 7-8). A systematic prevention of

100-

COCO

CC

E

CDN

CD.Q

O

ÇZ

Ecc

I

<frI

o

Eo

99-

98-

97-

96-

95-

unmodified 1 % CH20

T T

2% CH20 3 5%CH20 5% CH20

Fig. 7-8: Influence of the modification of Raney nickel by formaldehyde on the

hydrogénation of l-bromo-4-nitrobenzene. Reaction conditions: 10 g

l-bromo-4-nitrobenzene, 90 g THF, 60°C. 0.5 MPa and 1500 rpm.

the dehalogenation was not observed. A loss of activity was observed, as it

is demonstrated in Figure 7-9 and Table 7-3.

Page 99

unmodified

• 1%CH20 modified

* 2% CH20 modified

3.5% CH20 modified

5% CH20 modified

0-1,: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

0 200 400 600 800 1000 1200 1400 1600 1800 2000

time / [min]

Fig. 7-9: Hydrogen uptake (from a 37.5 ml storage vessel thermostated at 25°C) of

modified Raney nickel. Reaction conditions: 10 g l-bromo-4-nitrobenzene,

90 g THF, 60°C. 0.5 MPa and 1500 rpm.

Table 7-3: Summary of the influence of formaldehyde modified Raney catalysts on the

initial hydrogénation rate and the selectivity towards l-bromo-4-

aminobenzene.

modification initial hydrog

rate

enation l-bromo-4-

aminoazobenzene

/[-] / [mmolAV'kg)] / [mass-%)]

unmodified 22.2 99.3

1% CH20 11.5 98.5

2% CH20 7.8 98.4

3.5%CH20 7.5 98.6

5% CH20 4.8 99.4

aniline

/ [mass-%>]

0.7

0.5

1.6

1.4

0.6

Page 100

7.3 Hydrogénation of levodione

7.3.1 General remarks

The low pressure hydrogénation of levodione ((6R)-2,2,6-trimethylcyclo-

hexa-l,4-dione) to actinol ((4R,6R)-4-hydroxy-2,2,6-trimethylcyclohexa-

none) is an important industrial process. Thereby, the R,R isomer is the

desired product [99, 100]. A scheme of the reaction is given in Figure 7-10.

y^° h2 r^V0 r^V0«

+

0A/R\ "H2 HO<T^R% HO* S^^rX

Fig. 7-10: Reaction scheme for the hydrogénation of levodione to R,R- or R,S-actinol.

The reaction was carried out in a 100 ml three-neck sulphonation flask

(Chapter 9.1.3) according the described procedure in Chapter 9.2.5. The

catalyst was washed three times with methanol (Chapter 9.2.8) before the

reaction was started. The reaction conditions were: 10 g levodione, 50 g

MeOH, 25°C, 0.11 MPa and 600 rpm.


To check a possible influence of a transfer limitation on the hydrogénation

rate, the hydrogen uptake was monitored for two different catalyst amounts

(4 and 6 g Raney nickel). The reaction conditions were: 10 g levodione, 50 g

methanol, 25°C, 0.11 MPa and 600 rpm. The initial hydrogen uptake was

0.071 bar/(min*g) (4 g catalyst) and 0.10 bar/(min*g) (6 g catalyst),

respectively. Thus, a gas-liquid transfer limitation for hydrogen can be

Page 101

excluded, as the rate is linearly proportional to the amount of the catalyst

used.

7.3.3 Influence of the modification on selectivity and hydrogénation

rate

The influence of a modification by formaldehyde on the selectivity towards

the desired isomer and the hydrogénation rate was investigated using

unmodified and differently modified Raney catalysts. The influence on the

hydrogen uptake is plotted in Figure 7-11. The formaldehyde modified

catalysts have less activity then the unmodified. The selectivity towards the

desired R,R-actinol is shown in Figure 7-12. A higher selectivity towards

CD.Q

~ 15-

cu.*:

ro

Q.

cu

o

T3

30 -,

25-

20-

10-

400

time/[min]

Fig. 7-11: Hydrogen uptake during the hydrogénation of levodione using differently

modifiedRaney nickel. Reaction conditions: 10 g levodione, 4 g catalyst, 50 g

MeOH, 25°C. 0.11 MPa and 600 rpm.

Page 102

[CHp]/[%]

Fig. 7-12: Selectivity of the hydrogénation of levodione using differently modified

Raney catalysts. Reaction conditions: 10 g levodione, 4 g catalyst, 50 g

MeOH, 25°C. 0.11 MPa and 600 rpm.

R,R-actinol was not observed. The reported selectivity of 80-85%

R,R-actinol [99] could not be confirmed, even not if fresh catalysts were

used. The distribution of R,R-actinol and R,S-actinol is about 1:1. The

possible isomerisation of the hydrogénation products, can be explained

either by the sodium hydroxide present in the catalyst or by residual acid in

the levodione. Both, acid as well as base, catalyse the isomerisation of the

product.

7.4 Discussion

The hydrogénation of other substances than nitriles was performed to find

other potential applications for formaldehyde treated nickel catalysts. These

systems were screened without a profound study of the reaction mechanism.

Page 103

The modification of nickel catalysts by formaldehyde has no benefit

with respect to a selective hydrogénation of crotonaldehyde. The desired

crotylalkohol is not preferentially produced using the modified catalyst. The

modification only lowers the hydrogénation rates and causes undesired

by-products formed by condensation reactions on the catalyst.

If a halogenated nitrobenzene is hydrogenated, a modification by

formaldehyde does not systematically inhibit the dehalogenation. However,

a decrease in activity was observed.

The hydrogénation of levodione using formaldehyde treated catalysts

does not improve the yield to the desired R,R-actinol. Only a decreased

activity was observed.

Summing up, until now no other hydrogénations than that of nitriles

were found for which a modification of the nickel catalysts by

formaldehyde has been beneficial.

Page 104

Chapter

Conclusions and outlook

The modification of Raney nickel by formaldehyde leads to higher

selectivities towards the primary amines, also in cases that aliphatic nitriles

are hydrogenated. Up to 50% of the formaldehyde used to modify the

catalyst were found as methanol in the modification solution. A possible

explanation for this behavior is that formaldehyde disproportionates at the

catalyst to methanol and chemisorbed carbon monoxide.

If amines have to be produced from nitriles at large scale, several points

have to be thoroughly considered to produce high selectivities:

• The reaction time until the hydrogénation is finished has to be known,

so that the reaction mixture can be cooled immediately after full

conversion (kinetically controlled product distribution) to avoid

disproportionation reactions, that lower the selectivity

(thermodynamically controlled product distribution).

Low hydrogen pressures favour the production of primary amines, thus

a hydrogénation in the gas-liquid transfer limitation for hydrogen (high

catalyst loadings) may be favorable.

The modification of catalysts by formaldehyde offers a new way to

increase the selectivity towards primary amines, especially in cases that

aromatic nitriles are hydrogenated. In order to find the optimal

modification concentration investigations have to be performed because

desired higher selectivities are also accompanied by undesired lower

reaction rates.

An optimal temperature has to be found, because high temperatures lead

to higher selectivities towards primary amines (higher activation energy

of the hydrogénation to the primary amine than of the hydrogénation to

the secondary amine), but disproportionation reactions leading to the

Page 105

8

product distribution determined by the thermodynamic equilibrium can

lower the selectivity at high temperatures [27].

Page 106

Chapter ^

Experimental

9.1 Apparatus

9.1.1 Description of the 500 ml steel hydrogenator

Hydrogénations of butylamine and crotonaldehyde were carried out in a

500 ml hardening vessel (built by F. Hoffmann-La Roche, Figure 9-1). The

apparatus is designed to withstand a maximum pressure of 10 MPa and a

maximal temperature of 200°C. The autoclave is equiped with a four-bladed

agitator that can be rotated at rates between 10 and 2000 rpm.

The reactor can be operated at temperatures between 20 and 200°C. The

heating is electrical and cooling is achieved using cooling water at 15°C.

The autoclave consists of a reaction vessel and a cap, both made of rust-

resistant steel (W. No. 1.4435). A silverseal and 8 screws ensure tightness.

The cap has 7 boreholes (Figure 9-2), and within a central bore for the

agitator shaft (1). Hydrogen is introduced from the side via the cap. Samples

can be taken via valve 4 (Nova Swiss).

Liquid or gaseous starting products are added via valve 7 (Nova Swiss).

The temperature is measured with a type PT100 sensor (range: -100 to

400°C, precision: ±0.33%) manufactured by Rotax. This sensor is

positioned in borehole 2. The temperature is controlled by cascade

regulation.

The pressure in the autoclave is measured with a piezometer PA-23 100

(range: 0.1 to 10 MPa, precision: ±0.05 MPa) manufactured by Rotax (8).

The pressure in the hydrogen storage vessel is measured using a piezometer

PA-23 200 (range: 0.1 to 20 MPa, precision ±0.1 MPa) also manufactured

by Rotax. A bursting disk (Sitec) is also installed (bursting pressure: 11 MPa

Page 107

degassing

sampling system

temperature measurement

V\/\/WVWV\/W\^ / /W\A NW\MVW\AVWWWWWWWVWWWW\

copper block with electrical heating and cooling

hydrogen storage vessel

09

Fig. 9-1 : Hydrogénation apparatus.

±10%) in the cap (9). The autoclave pressure can be released by a valve (3,

Sitec). Nitrogen is fed to the autoclave via valve 6 (Nova Swiss). The

remaining holes (5 and 7) were not used and were therefore fitted with blind

flanges.

Page 108

1 agitator shaft

2 temperature measurement

3 pressure release

4 valve for sampling5 blind flange6 nitrogen inlet

7 blind flange8 pressure measurement

9 bursting disk

Fig. 9-2: Hydrogénation apparatus: Top view of the cap.

9.1.2 200 ml glass hydrogenator

A 200 ml glass hydrogénation vessel from Bilchi AG (max. pressure 1 MPa,

max. temperature 200°Cj was mounted on an apparatus containing a Bilchi

Cyclo 075 stirrer (max. speed 3000 rpm), a PA-23 10 piezometer (Rotax,

range 0.1 to 1 MPa), a PT 100 B (Rotax, -100 to 400°C, ±0.33%) and a

bursting disk (Sitek, 1 MPa bursting pressure). Two hydrogen storages,

thermostated at 25°C with a Lauda E 100 thermostat, were used (5 ml and

50 ml, total hydrogen storage volume inclusive pipes: 73.5 ml) during the

reaction. The glass hydrogenator was thermostated with a Lauda E 200

thermostate. The reaction data were stored in aEurotherm Chessel 4100 G

9.1.3 100 ml low pressure hydrogénation apparatus

A three-neck sulphonation flask, equiped with a glass thermometer and a

gas inlet (inlet of argon, hydrogen and pressure release on the same pipe)

was mounted on a stirrer system (the third hole was fitted with a blind

flange). A Julabo PC/4 thermostat was used to control the reaction

Page 109

temperature. The hydrogen uptake was monitored by a W+W recorder 312

from Kontron AG attached to a piezometer in the storage vessel. A bursting

disk (Sitek) was installed with a bursting pressure of 0.07 MPa.

9.1.4 Lab Shaker

A Lab Shaker from KühnerAG with 35 ml hydrogénation reactors (built by

F. Hoffmann-La Roche) was used to perform the reversibility experiments.

These autoclaves were equiped with a bursting disk (30 MPa) and a valve.

A shaking frequency of 280 min"1 on 25 mm was applied.

9.1.5 Modification and washing apparatus

Modification and washing procedures were carried out in a three-neck

sulphonation flask, equiped with a nitrogen intake and an outlet connected

to a bubble counter. A glass stirrer driven by a motor was installed in the

central grinding. If the catalyst was modified, a dropping funnel was

installed in the last vent, otherwise this vent was provided with a blind plug.

9.1.6 Gas Chromatograph

An HP 6890 Series GC System with an HP 7683 Series Injector

autosampler was used. A flame ionisation detector was used for normal

measurements. An HP 5973 Mass Selective Detector was used for gas

chromatography with mass spectroscopic detection. HP ChemStation

software, version Rev.A. 06.03 [509] was used to control and analyse the

measurements. By aid of this program an analysis method was developed

[101-103].

Page 110

9.2 Methods

9.2.1 Hydrogénation of butyronitrile

240 g (3 mol) butyronitrile were placed in the 500 ml steel autoclave

(Chapter 9.1.1) before the catalyst (if Raney nickel was used, the catalyst

was added as wet paste) was added. The autoclave was sealed and purged

three times with 0.55 MPa nitrogen. The reactor was then heated while the

content was stirred slowly (600 rpm). Once the reaction temperature had

been reached, the stirrer was turned off and hydrogen was pressed into the

autoclave. The reaction was initiated by turning on the stirrer. Samples were

taken during the hydrogénation (Chapter 9.2.6).

After the reaction was completed, the reaction vessel was cooled to

room temperature, the pressure released and the vessel purged three times

with 0.55 MPa nitrogen. The reaction mixture was sucked off by a vacuum

pump and then filtered. The autoclave and the filtering apparatus were

purged with methanol. To clean the autoclave, the vessel was filled with

methanol which then was boiled at 80°C for 15 min.

The Raney nickel was disposed into a special nickel waste jar and the

reaction mixture in a solvent waste jar.

9.2.2 Reversibility experiments in the 35 ml screening autoclave

20 g butylamine (0.27 mol) were placed in a 35 ml screening hydrogenator

before 2 g of wet Raney nickel catalyst were added. The hydrogenator was

then closed and purged three times with 4.2 barg nitrogen. Ammonia or

hydrogen were then added via the valve on the autoclave. The autoclave was

placed in the Lab Shaker and heated to reaction temperature while being

shaken at 280 min"1.

After a defined reaction time the screening hydrogenator was placed in

a cooling block (copper block cooled by water), the pressure was released,

the reactors were again purged three times with 4.2 barg nitrogen and then

Page ill

opened. The content of the reactors was flushed out with methanol. The

reaction mixture was then filtered and analysed.

9.2.3 Hydrogénation of crotonaldehyde

40 g (0.57 mol) butyronitrile and 200 g ethanol were placed in the 500 ml

steel autoclave (Chapter 9.1.1) before the catalyst (ifRaney nickel was used,

the catalyst was added as wet paste) was added. The autoclave was sealed

and purged three times with 0.55 MPa nitrogen. The reactor was then heated

while the content was stirred slowly (600 rpm). Once the reaction

temperature had been reached, the stirrer was turned off and hydrogen was

pressed into the autoclave. The reaction was initiated by turning on the

stirrer. Samples were taken during the hydrogénation (Chapter 9.2.6).

After completion of the reaction, the reaction vessel was cooled to room

temperature, the pressure released and the vessel purged three times with

0.55 MPa nitrogen. The reaction mixture was sucked off with a vacuum


purged with ethanol. To clean the autoclave, the vessel was filled with

ethanol which then was boiled at 80 °C for 15 min.



9.2.4 Hydrogénation of l-bromo-4-nitrobenzene

10 g (50 mmol) l-bromo-4-nitrobenzene and 100 g tetrahydrofurane were

placed in the 200 ml glass autoclave (Chapter 9.1.2) before the catalyst was

added as a wet paste. The autoclave was sealed and purged three times with

0.3 MPa nitrogen. The reactor was then heated while the contents were

stirred slowly (600 rpm). Once the reaction temperature had been reached,

the stirrer was turned off and hydrogen was pressed into the autoclave. The

reaction was initiated by turning on the stirrer.

After completion of the reaction, the reaction vessel was cooled to room

temperature, the pressure released and the vessel purged three times with

Page 112

0.3 MPa nitrogen. The reaction mixture was sucked off with a vacuum


finally purged with tetrahydrofurane.



9.2.5 Hydrogénation of levodione

10 g (65 mmol) levodione ((6R)-2,2,6-trimethylcyclohexa-l,4-dione) and

50 g methanol were placed in a 100 ml three-neck sulphonation flask

(Chapter 9.1.3) before the catalyst was added as a wet paste. The flask was

installed on the hydrogénation equipment, evacuated and purged three times

with 0.11 MPa argon. The reactor was then heated while the contents were

stirred slowly (600 rpm). Once the reaction temperature had been reached,

the stirrer was turned off and hydrogen was pressed into the autoclave.

Then, the reaction was started by turning the stirrer on again.

After the reaction was completed, the reaction vessel was cooled to

room temperature, the pressure released and the vessel evacuated and

purged three times with 0.11 MPa argon. The reaction mixture was sucked

off with a vacuum pump and then filtered. To clean the autoclave and the fil¬

tering apparatus they were purged with methanol.



9.2.6 Description of the sampling procedure

Samples were taken with a steel capillary with 1 mm inner diameter and a

2 jim frit (dead volume: 0.5 ml). Because sometimes the reaction

temperatures were higher than the boiling temperatures of the substances,

the samples were collected in a test tube with solvent (methanol or ethanol in

the case of butyronitrile or crotonaldehyde, respectively). These test tubes

were cooled in a Dewar vessel containing dry ice and ethanol. Initially, 2 ml

ofthe reaction mixture were rejected before 1 ml was collected and analysed.

Page 113

9.2.7 Neutralisation with water

50 g Raney nickel were suspended in 100 ml distilled water in a three-neck

sulphonation flask. The flask was purged three times with argon before the

suspension was stirred for 30 min. The catalyst was then allowed to settle

before the solvent was decanted. This procedure was repeated three times

with distilled water. The neutralised catalyst was stored in distilled water.

9.2.8 Neutralisation with methanol




before the solvent was decanted. This procedure was first carried out once

with distilled water and then three times with methanol. The neutralised

catalyst was stored in methanol.

9.2.9 Neutralisation with tetrahydrofurane




before the solvent was decanted. This procedure was first carried out once

with distilled water and then three times with tetrahydrofurane. The

neutralised catalyst was stored in tetrahydrofurane.

9.2.10 Modification with formaldehyde


sulphonation flask. The flask was purged three times with argon before

13 ml of a solution of 35% formaldehyde were added slowly. The


before the solvent was decanted. The catalyst was washed three times with

Page 114

30 ml distilled water (in accordance with Chapter 9.2.7), and the modified

catalyst was stored in methanol.

9.3 Analytics

9.3.1 Determination of butyronitrile, butylamine, dibutylamine and

dibutylimine with a GC method using an internal standard

GC conditions

apparatus HP 6890 gas Chromatograph with

split injector and FID

HP 7863 autosampler

column stationary phase Rtx-5 Amine

length x ID 30 m x 0.32 mm.,

film 1.0 jim

column material 5% diphenyl- 95% dimethyl polysiloxane

manufacturer Restek Corporati;on

carrier gas helium pressure 75kPa

total flow 112 ml/min

split ratio 50:1

column temperature 65°C(CImin), 3°/min, 80°C(0min), 20°C/min;

280°C(0 min)

injector temperature 250°C

detector temperature 250°C

injection volume ljLLl

Sample preparation

ISTD solution 5 gn-caprylic acid methyl ester were diluted in 1 ]

methanol.

Page 115

calibration solution

sample solution

5-10 mg of reference substances were weighted in

a sample vial before 1 ml internal standard

solution was added.

15 jil of the collected sample were diluted in 1 ml

internal standard solution.

analysis time

retention times butylamine

butyronitrile

dibutylimine

dibutylamine

ISTD

tributylamine

15.0 min

2.5 min

2.9 min

7.4 min

8.0 min

10.1 min

10.6 min

Remarks

Calibrations were performed using four different concentrations of the

substances to be analysed.

9.3.2 Determination of crotonaldehyde, crotylalkohol, butanal and

butanol with a GC method using an internal standard

GC conditions



HP 7863 autosampler

column stationary phase

length x ID

column material

manufacturer

Stabilwax

30 m x 0.32 mm, film 0.25 jim

Carbowax-PEG

Restek Corporation

Page 116

carrier gas helium pressure 75kPa


split ratio 50:1

column temperature 40°C(5min), 37min,

240°C(3.6min)

70°C(0min), 15°C/min



injection volume l|il

Sample preparation

ISTD solution 5 g n-caprylic acid metlîyl ester were diluted in 1


sample solution

ethanol.

5-10 mg of reference substances were weighted in

a sample vial before 1 ml internal standard

solution was added.

100 jil were diluted in 1 ml internal standard

solution.

analysis time

retention times butanal

crotonaldehyde E

crotonaldehyde Z

butanol

crotylalkohol E

crotylalkohol Z

ISTD

30.0 min

2.4 min

4.9 min

5.0 min

8.5 min

11.7 min

12.9 min

17.8 min

Remarks

Crotonaldehyde and crotylalkohol were calibrated and measured as the sum

of the E and Z isomers. Four samples of different concentrations were used

for calibration.

Page 117

9.3.3 Determination of l-bromo-4-nitrobenzene, l-bromo-4-

aminobenzene and aniline

GC conditions



HP 7863 autosampler

column stationary phase DB-5HT

length x ID 15 m x 0.25 mm, film 0.1 jim

column material 5% diphenyl - 95% dimethyl polysiloxane

manufacturer J&WScientific

carrier gas helium pressure 57 kPa


split ratio 25:1

column temperature 60°C(0min), 107min, 250°C(0min)



injection volume l|il

Sample preparation

sample solution 500 jil of the filtrated reaction mixture wen

diluted with 500 jllI THF.

analysis time 19.0 min

retention times aniline 1.9 min

1 -bromo-4-nitrobenzene

1 -bromo-4-aminobenzene

2.9 min

5.2 min

Page 118

9.3.4 Determination of levodione and actinol

GC conditions



HP 7863 autosampler


length x ID

column material

manufacturer

HP-5

30 m x 0.32 mm, film 0.25 jim

5% diphenyl - 95% dimethyl polysiloxane

Hewlett Packard

carrier gas helium pressure

total flow

split ratio

79 ml/min

50:1

column temperature

injector temperature

detector temperature

injection volume

50°C(0min), 107min, 150°C(3min)

250°C

300°C

ljLLl

Sample preparation

sample solution 100 jil of the filtrated reaction solution were

diluted with 900 jil methanol

analysis time

retention times levodione

(R,S)-actinol

(R,R)-actinol

13.0 min

8.8 min

9.8 min

9.9 min

Page 119

9.3.5 Methanol determination in aqueous medium with a headspace GC

method using an external standard

GC conditions

apparatus Agilent 6890N gas Chromatograph with FID

PE HS 40 XL headspace autosampler


length x ID

column material

manufacturer

DB-1

30 m x 0.32 mm, film 5 jim

100% dimethyl polysiloxane

J&WScientific

carrier gas nitrogen pressure 20 ps

splitless

column temperature 40°C(3min), 107min, 150°C(2min)

transfer temperature 100°C



injection time 0.08s

Sample preparation


sample solution

a solution of 100 jil MeOH in 1000 ml water was

prepared as ESTD, 1 ml of this solution was

placed in the HS autosampler.

the aqueous modification solution was diluted

1:10 with water before 100 mg were again diluted

in 1 ml water and then placed in the HS

autosampler.

analysis time

retention times methanol

16.0 min

1.6 min

Page 120

9.3.6 Formaldehyde determination in aqueous medium with an HPLC

method using an external standard

HPLC conditions

apparatus HP 1050 HPLC with UV detector


length x ID

manufacturer

Supelcosil LC-18

250 mm x 4.6 mm, film 5 jim

Supelco

solvent acetonitrile : water 60 : 40

flow 1.5ml/min

detection wavelength

injection volume

360 nm

2 |il

Sample preparation

derivatization solution

derivatization


2.37 g (8 mmol) 2,4-dinitrophenylhydrazine were

dissolved in 100 ml THF.

5 ml of the derivatization solution were tared,

before 30 mg of the sample were added. 1 drop of

cone, hydrochloric acid was added and the sample

was stored for 1 hour at 50°C. 1 ml of the warm

derivatized solution was diluted with 10 ml

acetonitrile and then analysed.

a solution containing 5% formaldehyde was

derivatized and then used as ESTD.

analysis time

retention times 2,4-dinitrophenylhydrazone

20.0 min

3.8 min

Page 121

9.3.7 Synthesis of dibutylimine as a standard for GC measurements

Butylamine condenses with butyraldehyde to produce the Schiff base

(Figure 9-3) [104, 105].

Fig. 9-3: Condensation of butylamine and butyraldehyde [104].

34.6 g (506 mmol) butyraldehyde in 56 g toluene were placed in a three-

neck sulphonation flask before 36.8 g (504 mmol) butylamine were added

slowly via a dropping funnel. The solution was heated to 66°C and water

was produced. After one hour the aqueous phase was separated off and the

organic phase was distilled.

The same conversion was also made on a smaller scale (calibration

samples for gas chromatographic analysis) and on a larger scale (to produce

azomethine as starting material for the hydrogénation), in both cases

without solvent and using an excess of butylamine. This conversion is

quantitative, no butyraldehyde was detected after the reaction.

The condensation of dibutylamine and butyraldehyde is slower than the

one of butylamine and butyraldehyde, but using the same conversion,

tributyleneamine could be produced. This was not done because

tributyleneamine was not detected during the hydrogénation of butyronitrile

in the reaction mixture.

Page 122

9.4 Identification of by-products

9.4.17V-Butylbutanamide

Initially, a reaction mixture, containing 4 mass-% TV-butylbutanamide was

concentrated and distilled under vacuum. The obtained sample was purified

by chromatography using ethyl acetate/light petroleum (50 : 50) as eluent to

yield the pure compound. A DPX 400 NMR spectrometer (400 MHz 1H, 75

MHz 13C) was used to measure the nuclear magnetic resonance spectra;

Ôh(CDC13) 0.92 (3H, t, .7=8.0, CH3), 0.95 (3H, t, .7=8.2, CH3), 1.34 (2H, m,

C772CH2CH2NH), 1.48 (2H, m, C772CH2NH), 1.66 (2H, sext, J=l .5

C772CH2CO), 2.15 (2H, t, .7=7.4, CH2CO), 3.25 (2H, q, .7=6.7, CH2NH),

5.67 (1H, s, NH); 5C(CDC13) 14.1 (2CH3), 19.6 (CH2CH2CO), 20.4

(CH2CH2CH2N), 32.1 (CH2CH2N), 39.2 (CH2CO), 39.6 (CH2N) and 173.4

(CO). A mass spectrum was measured using a GC-MS system (5973 Mass

Selective Detector); m/z 143 (M+, 19%), 128 (C7H14NO+, 23), 115 (29), 101

(25), 100 (C5H10NO+, 57), 88 (20), 86 (C4H8NO+, 19), 73 (52), 71 (C4H70+,

90), 57 (C4H9+, 35), 44 (62), 43 (C3H7+, 95), 41 (60) and 30 (C2H6+, 100).

These results are in accordance with published data [106].

9.4.2 7V,7V-Dibutylbutyramidine

The reaction mixture containing about 7 mass-% AyV-dibutylbutyramidine

was concentrated to a small volume and then distilled in vacuo. A sample of

15 mass-% TV-butylbutanamide and 85 mass-% AyV-dibutylbutyramidine

was obtained. An AV 500 NMR spectrometer (500 MHz 1H, 125 MHz 13C,36 MHz 14N (ref. nitromethane)) was used to measure the nuclear magnetic

resonance spectra; 5H(CDC13) 0.94 (6H, t, .7=7.4, CH3CH2CH2CH2N), 1.09

(3H, t, .7=7.4 CH3CH2CH2C), 1.37 (4H, m, C772CH2CH2N), 1.63 (6H, m,

C772CH2N, C772CH2C), 2.34 (2H, m, CH2C), 3.19 (4H, t, .7=7.5 CH2N),

7.29 (1H, s, NH); 5C(CDC13) 13.6 (2 CH3CH2CH2CH2N), 14.1

(CH3CH2CH2C) 19.7 (2 CH2CH2CH2N), 20.0 (2 CH2CH2N), 27.5

(CH2CH2C), 31.8 (CH2C), 43.3 (2 CH2N), 137.6 (CNH); ôN(CDCl3, 300K)

Page 123

-165, line width 90, -280, line width 75; ÔN(CDC13, 333K) -230 line width

175; ôN(TFA/dioxane, 300K) -260, line width 125. An NMR literature

reference for aromatic amidines is given in [107]. A mass spectrum was

measured using a GC-MS system (5973 Mass Selective Detector); m/z 198

(M+, 20%), 183 (CnH23N2+, 24), 169 (C10H21N2+, 29), 155 (C9H19N2+, 17),

141 (C8H17N2+, 17), 127 (C8H17N+, 45), 113 (16), 99 (15), 85 (15), 84 (78),

72 (C4H10N+, 26), 70 (C4H8N+, 100), 57 (C4H9+, 17), 43 (C3H7+, 24), 41

(33)and29(C2H5+,23).

9.4.3 1,1-Diethoxybutane

A mass spectrum was measured using a GC-MS system (5973 Mass

Selective Detector); m/z 103 (C5Hn02+, 100%), 101 (C6H130+, 85), 75 (50),

73 (M2+, 40), 55 (48), 47 (38), 43 (C3H70+, 16) and 29 (C2H50+, 14).

9.5 Characterisation of the catalyst

9.5.1 The reduction potential

The reduction potential of Raney nickel was measured with two different

combined gold electrodes with an integrated Ag/AgCl reference, model

6.0413.100, manufactured by Metrohm. The measurement was monitored

by a Metrohm pH-meter 691. The electrodes were calibrated with Metrohm

redox standard 6.2306.020 (U=250±5 mV). When the redox potential was

measured, the electrodes were placed directly on the solid catalyst. The

values obtained were converted into the standard reduction potentials

relative to the Pt/H2-electrode.

Page 124

9.5.2 Adsorption of an indicator

0.01-0.02 g of wet Raney nickel were placed in a 100 ml volumetric flask

before 100 ml of a 10"4 M 4-aminoazobenzene solution were added. The

solution was allowed to stand for 4 weeks, before the concentration of

4-aminoazobenzene was measured with an UVIKON 720 LC UV-VIS

absorption spectrometer. From the difference in concentration the amount of

the adsorbed indicator was calculated.

9.5.3 Dissolution in acidic medium

1 g of wet Raney nickel was placed in a three-neck sulphonation flask

before 50 g water were added. With a dropping funnel 32 ml of hydrochloric

acid (35%) were slowly added while the content of the flask was stined with

a magnetic stirrer. The released gas was measured with a measuring cylinder

that was filled with water inversely placed in a water containing vessel.

9.6 Chemicals

The chemicals used, their suppliers and purity grades are listed in Table 9-1.

Table 9-1 : Purity values and suppliers of the chemicals used

Substance Supplier Purity grade

4-aminoazobenzene Merck for synthesis > 98% (HC104)

aniline Fluka puriss. p. a. > 99.5% (GC)

l,2-bis(2-hydroxyethyl- F. Hoffmann-La Roche Lot No. 28611

thio)ethane (EDS)

1 -bromo-4-nitrobenzene Fluka puriss. p. a. > 98% (GC)

butanol Fluka puriss. p. a. > 99.5% (GC)

Page 125

Table 9-1 : Purity values and suppliers of the chemicals used

Substance Supplier Purity grade

butyraldehyde Fluka purum > 97% (GC)

butylamine Fluka purum > 98% (GC)

butyronitrile Fluka purum > 99% (GC)

n-caprylic acid methyl ester Fluka puriss > 99% (GC)

crotonaldehyde (E + Z) Fluka purum > 98% (GC)

crotylalkohol (E + Z) Fluka purum > 97% (GC)

dibutylamine Fluka puriss. > 99% (GC)

ethanol Merck p. a.

formaldehyde solution Merck extra pure 35%

formaldehyde solution Fluka puriss. p. a. ACS free of acids

hydrochloric acid 25% Merck GR for analysis

levodione F. Hoffmann-La Roche Lot No. 32716-a8

methanol Merck p. a.

methanol F. Hoffmann-La Roche tech.

nickel-on-carrier Engelhard Ni 1404 P, Lot H-99

pyridine Fluka p. a. > 99.8% (GC)

tetrahydrofurane Fluka puriss. p. a. > 99.5% (GC)

toluene Fluka puriss. p. a. > 99.5% (GC)

tributylamine Fluka puriss. p. a. > 99% (GC)

argon Carbagas 99.995%

nitrogen Carbagas 99.995%

hydrogen Carbagas 99.995%

Raney nickel Degussa-Hülls AG Type B113 Z, Batch 20018989

Page 126

9.7 Calculation of selectivity and reaction rates

In every experiment samples were taken after 10, 20, 30, 40, 50, 60, 75,

90, 105, 120, 150, 180 min and, if necessary, every further hour. With aid of

these samples reaction profiles were drawn (an example is given in

Figure 9-4) and the selectivity and the initial reaction rates were calculated.

100-

„80-

t/5to

ce60-

's 4°-

CDÜ

OÜ

20-

• m

— butylamine

• butyronitrile* dibutylimine

— dibutylamine

Br em.__

gfTTTTV

-

0 50 100 150 200 250 300 350 400

time / [min]

Fig. 9-4: Typical reaction profile for the hydrogénation of butyronitrile. Reaction

conditions: 240 g BN, 3.75 g RaNi, 100°C, 1 MPa overall pressure,

1000 rpm.

The selectivities of the substances were calculated with the following

equation (Eq. 9.1):

selectivity(x) = mass of component x / total mass Eq. 9.1

Page 127

Initially, to calculate the initial reaction rates, polynomials of second order

were fitted to the data in a range of 0 to 50% butyronitrile conversion. The

reaction rates were then obtained from the first derivatisation of the fitted

function at the point of zero butyronitrile conversion (origin). The obtained

values /[mass-%/min] were converted to molar rates, and these were divided

by the mass of charged catalyst in order to obtain the molar rates per mass

catalyst and second (/[mmol/kg*s]).

9.8 Error analysis

9.8.1 Precision of a gas chromatographic analysis

The precision of an analysis by gas chromatography was measured with a

calibration sample. Five measurements were carried out (Table 9-2). The

relative standard deviation was 1.2% in the worst case.

Table 9-2: Standard deviation (g) and relative standard deviation (rel. g) of a gas chro¬

matographic measurement.

substance

butylamine

butyronitrile

dibutylamine

tributylamine

mean G rel. G

/ [mass-%)] /[mass-%)] /[%]

18.55 0.15 0.81

40.48 0.44 1.09

30.17 0.36 1.19

10.81 0.09 0.84

Page 128

9.8.2 Precision of a sample analysed with gas chromatography

64.14 0.27 0.42

20.87 0.49 2.33

6.52 0.14 2.14

8.46 0.21 2.52

The precision of the sampling procedure with gas chromatography was

measured in experiment 5 (conditions: 240 g BN, 15g methanol washed

RaNi, 100°C, 1 MPa, 1000 rpm) after a reaction time of 120 min. Five

samples were taken and analysed (Table 9-3).

Table 9-3: Standard deviation (g) and relative standard deviation (rel. g) of different

samples.

substances mean g rel. g

/ [mass-%] / [mass-%] / [%]

butylamine

butyronitrile

dibutylamine

dibutylimine

The relative standard deviation was 2.6% in the worst case. It must be noted

that the sampling procedure for five samples lasted 3 min and that during

this time the reaction was running. This value is in an expected range and is

comparable to literature values of about 2% [27].

Concerning the accuracy of the measurement the following has to be

mentioned: To calculate the relative concentrations, all compounds in the

sample were determined in [mg] and divided by the total mass. Compounds

not calibrated were considered using a relative response factor of 1.0

relative to the internal standard. A systematic enor is therefore made when

calculating the total mass and of course in the concentrations of the

components. This enor depends on the concentration of substances that are

not calibrated. Two not calibrated compounds were detected: dibutylamide

and tributylamidine.

Page 129

9.8.3 Precision of the selectivity and the reaction rates

The precision of the selectivity and the reaction rates was measured by

repeating an experiment twice (Table 9-4 and Table 9-5). Different

calibrations of the gas Chromatograph were used and the time periods

between the experiments were several months. The conditions were: 240 g

BN, 15 g RaNi, 100°C, 1 MPa hydrogen, 1000 rpm.

Table 9-4: Precision of the selectivity of the hydrogénation of butyronitrile. Reaction

conditions: 240 g butyronitrile, 15 g RaNi, 100°C, 1 MPa, and 1000 rpm.

BA DBA Tamidine max. DBI

/ [mass-%)] / [mass-%)] / [mass-%)] / [mass-%)]

exp 1 90.98 8.50 0.53 10.80

exp 2 90.73 8.84 0.43 9.73

exp 3 91.13 8.07 0.61 10.46

mean 90.95 8.47 0.52 10.33

standard deviation 0.20 0.38 0.09 0.55

Table 9-5 : Precision ofthe initial reaction rates. Reaction conditions: 240 g butyronitrile,

15 g RaNi, 100°C, 1 MPa, and 1000 rpm.

d[BN]/dt d[BA]/dt d[DBA]/dt d[DBI]/dt

/ [mmol/kg*s] / [mmol/kg*s] / [mmol/kg*s] / [mmol/kg*s]

exp 1 65.7 46.6 1.40 7.97

exp 2 60.7 43.6 1.16 7.35

exp 3 67.0 48.7 0.97 8.17

mean 64.5 46.3 1.18 7.83

standard deviation 3.3 2.5 0.21 0.43

Page 130

Chapter

Literature

[I] O. G Degischer, F. Roessler, US Patent 2001004672, 2001.

[2] S. Nishimura, in "Handbook of Heterogeneous Catalytic

Hydrogénation for Organic Synthesis", John Wiley & Sons, New

York, 2001, pp. 254-285.

[3] P. N. Rylander, in "Catalytic Hydrogénation in Organic Synthesis",

Academic Press, New York, 1979, pp. 138-152.

[4] P. N. Rylander, in "Hydrogénation Methods", Academic Press, New

York, 1985, pp. 94-103.

[5] E. Müller, O. Bayer, in "Methoden der organischen Chemie",

Thieme Verlag, Stuttgard, 1980, IV/lc, Reduktion Teil 1, pp.

110-144.

[6] F. Zymalkowski, in "Katalytische Hydrierungen im Organisch-

Chemischen Laboratorium", Enke Verlag, Stuttgart, 1965, pp.

237-257.

[7] R. Kuhn, W. Kirschenlohr,Liebigs Ann. Chem., 1956, 600, 115-125.

[8] R. Kuhn, H. Grassner, Liebigs Ann. Chem., 1958, 612, 55-64.

[9] E. Möltgen, P Tinapp, Liebigs Ann. Chem., 1979, 1952-1959.

[10] J. Braun, G Blessing, F. Zobel, Ber, 1923, 36, 1988-2001.

[II] H. Greenfield, Ind. Eng. Chem. Prod. Res. Dev, 1967, 6, 142-144.

[12] J. Volf, J. Pasek, Stud. Surf. Sei. Catal., 1986, 27, 105-144.

[13] J. L. Dallons, A. van Gysel, G. Jannes, Chem. Ind. (Dekker), 1992,

47,93-104.

[14] M. J. F. M. Verhaak, A. J. van Dillen, J. W. Geus, J. Catal., 1993,

143, 187-200.

[15] M. J. F. M. Verhaak, A. J. van Dillen, J. W. Geus, Appl. Catal. A,

1993,705,251-269.

Page 131

[16] M. J. F. M. Verhaak, A. J. van Dillen, J. W. Geus, Catal. Lett., 1994,

26,37-53.

[17] Y. Y. Huang, W. M. H. Sachtler, J. Phys. Chem. B, 1998, 102,

6558-6565.

[18] Y. Y. Huang, W. M. H. Sachtler, Appl. Catal. A, 1999,182, 365-378.

[19] Y. Y. Huang, W. M. H. Sachtler, J. Catal., 1999,184, 247-261.

[20] Y. Y. Huang, W. M. H. Sachtler, J. Catal., 1999, 7 88, 215-225.

[21] Y. Y. Huang, W. M. H. Sachtler, J. Catal, 2000,190, 69-1A.

[22] Y. Y. Huang, W. M. H. Sachtler, Stud. Surf. Sei. Catal, 2000,130 A,

527-532.

[23] C. V Rode, M. Arai, M. Shirai, Y. Nishiyama, Appl. Catal. A, 1997,

148, 405-413.

[24] B. Coq, D. Ticht, S. Ribet, J. Catal, 2000,189, 117-128.

[25] M. Besson, J. M. Bonnier, M. Joucla, D. Djaouadi, Bull. Soc. Chim.

Fr, 1990, 727, 5-19.

[26] C. de Bellefon, P. Fouilloux, Catal. Rev Sei. Eng, 1994, 36,

459-506.

[27] O. G. Degischer, Dissertation Nr. 14012, ETH Zürich, 2001.

[28] F. M. Cabello, D. Tichit, B. Coq, A. Vaccari, N. T. Dung, J. Catal,

1997, 7<57, 142-152.

[29] H. R. Christen, F. Vögtle, in "Organische Chemie, von den

Grundlagen zur Forschung", Otto Salle Verlag, Frankfurt, 1992, 1,

pp. 561-830.

[30] S. N. Thomas-Pryor, T. A. Manz, Z. Liu, T. A. Koch, S. K.

Sengupta, W. N. Delgass, Chem. Ind. (Dekker), 1998, 75, 195-206.

[31] T. A. Johnson, US Patent 5869653, 1999.

[32] M. Besson, D. Djaouadi, J. M. Bonnier, S. Hamar-Thibault, M.

Joucla, Stud. Surf Sei. Catal, 1991, 59, 113-120.

[33] O. Nicodemus, W. Schmidt, DE Patent 510439, 1930.

[34] P. Herold, K. Smeykal, DE Patent 626923, 1936.

[35] W. Reppe, E. Bauer, DE Patent 741683, 1943.

[36] H. Raab, DE Patent 738448, 1943.

[37] E. Hädicke, W. Schlenk, Liebigs Ann. Chem, 1972, 764, 103-111.

Page 132

[38] A. Baiker, W Caprez, W L. Holstein, Ind. Eng. Chem. Prod. Res.

Dev, 1983, 22, 217-225.

[39] J. F. Olin, T. E. Deger, US Patent 2192523, 1940.

[40] S. Nishimoto, B. Ohtani, T. Yoshikawa, T. Kagiya, J. Am. Chem.

Soc,1983, 705,7180-7182.

[41] M. J. F. M. Verhaak, A. J. van Dillen, J. W. Geus, Appl. Catal. A,

1994,109, 263-275.

[42] P. Sabatier, in "Die Katalyse in der Organischen Chemie",

Akademische Verlagsgemeinschaft, Leipzig, 1927.

[43] M. Raney, US Patent 1563587, 1925.

[44] M. Raney, US Patent 1628190, 1927.

[45] P Fouilloux, Appl. Catal, 1983, 8, 1-42.

[46] S. Nishimura, in "Handbook of Heterogeneous Catalytic

Hydrogénation for Organic Synthesis", John Wiley & Sons, New

York, 2001, pp. 7-19.

[47] K. Weissermel, H. J. Arpe, in "Industrial Organic Chemistry", VCH,

1993, pp. 64-72.

[48] G Ertl, H. Knözinger, J. Weitkamp, in "Handbook of Heterogeneous

Catalysis", VCH, 1997.

[49] R. L. Augustine, in "Heterogeneous Catalysis for the Synthetic

Chemist", Dekker, New York, 1996, pp. 473-510.

[50] P. Marion, US Patent 2001027257, 2001.

[51] G A. Martin, P. Fouilloux, J. Catal, 1975, 38, 231-237.

[52] D. Djaouadi, M. Besson, J. Jenck, P. Fouilloux, Chem. Ind.

(Dekker), 1995, 62, 423-429.

[53] J. R. Kosak, US Patent 3546297, 1970.

[54] J. F. Mais, K. C. Paetz, H. Fiege, H. U. Blank, D. Brueck, W. Mehl,

DE Patent 19604988, 1998.

[55] J. L. Dallons, G Jannes, B. Delmon, Stud. Surf. Sei. Catal, 1988, 41,

115-121.

[56] H. Lei, Z. Song, D. Tan, X. Bao, X. Mu, B. Zong, E. Min, Appl.

Catal. A, 2001, 214, 69-76.

[57] G Wegener, E. Waldau, B. Pennemann, B. Temme, H. Warlimont,

U. Kühn, DE Patent 19753501, 1999.

Page 133

[58] S. D. Mikhailenko, T. A. Khodareva, E. V. Leongardt,

A. I. Lyashenko, A. B. Fasman,J. Catal, 1993,141, 688-699.

[59] J. R. Kosak, US Patent 3145231, 1964.

[60] GB Patent 919273, 1963.

[61] P. Baumeister, W. Schener, US Patent 4960936, 1990.

[62] M. Studer, P. Baumeister, US Patent 6096924, 2000.

[63] O. G. Degischer, Laboruntersuchung zur Hydrierung von Pynitril,

VFCR, F. Hoffmann-La Roche, 2000.

[64] R. C. Balk, Benzonitril Hydrierung, VFCR, F. Hoffmann-La Roche,

2000.

[65] D. E. Gardin, G A. Somorjai, J. Phys. Chem., 1992, 96, 9424-9431.

[66] P D. Ditlevsen, D. E. Gardin, M. A. van Hove, G A. Somorjai,

Langmuir, 1993, 9, 1500-1503.

[67] V Bustos, M. V Gargiulo, J. L. Sales, R. O. Unac, G. Zgrablich,

Langmuir, 1997,13, 4301-4304.

[68] B. Bigot, F. Delbecq, V H. Peuch, Langmuir, 1995, 77, 3828-3844.

[69] B. Bigot, F. Delbecq, A. Milet, V H. Peuch, J. Catal, 1996, 159,

383-393.

[70] G. Blyholder, L. D. Neff, J. Catal, 1963, 2, 138-144.

[71] J. B. Butt, C. L. M. Joyal, C. E. Megiris, Chem. Ind. (Dekker), 1987,

30,3-37.

[72] C. H. Bartholomew, Appl. Catal. A, 2001, 272, 17-60.

[73] V Penchev, N. Davidova, V Kanazirev, M. Minchev, Y. Neinska,

Adv. Chem. Ser, 1973, 727, 461.

[74] J. G. McCarty, H. Wise, J. Catal, 1979, 57, 406-416.

[75] A. J. H. M. Kock, P. K. de Bokx, E. Boellaard, W. Klop, J. W. Geus,

J. Catal, 1985,9<5,468-480.

[76] F. Hochard, H. Jobic, J. Massardier, A. J. Renouprez, J. Mol. Catal,

1995,95,165-172.

[77] N. T. Dung, D. Tichit, B. H. Chiche, B. Coq, Appl. Catal, 1998,

169, 179-187.

[78] H. Knoezinger, in "Acid-Base Catalysis", VCH Verlagsge¬

meinschaft, Weinheim, 1988, pp. 147-167.

[79] M. Bodenstein, Z. Phys. Chem., 1913, 85, 329-397.

Page 134

[80] A. Chojecki, in "Selective hydrogénation of butyronitrile over

Raney-metals", Technische Universität München, 2004.

[81] O. G. Degischer, Pynitrilhydrierung Katalysatorscreening, VFCR,

F. Hoffmann-La Roche, 1999.

[82] L. J. Richter, W. Ho, J. Chem. Phys., 1985, 83, 5, 2165-2169.

[83] R. H. Newton, B. F. Dodge, J. Am. Chem. Soc, 1933, 55,

4747-4759.

[84] J. T. Yates, D. W. Goodman, J. Chem. Phys., 1980, 73, 10,

5371-5375.

[85] A. Bandara, S. Katano, J. Kubota, K. Onda, A. Wada, K. Domen,

C. Hirose, Chem. Phys. Lett, 1998, 290, 261-267.

[86] A. Bandara, S. Dobashi, J. Kubota, K. Onda, A. Wada, K. Domen,

C. Hirose, S. S. Kano, Surf Sei., 1997, 387, 312-319.

[87] A. Bandara, S.S. Kano, K. Onda, S. Katano, J. Kubota, K. Domen,

C. Hirose, A. Wada, Bull. Chem. Soc. Jpn., 2002, 75, 1125-1132.

[88] P. Claus, S. Schimpf, R. Schödel, P. Kraak, W. Mörke, D. Hönicke,

Appl. Catal. A, 1997, 7<55, 429-441.

[89] T. B. L. W. Marinelli, S. Nabuurs, V Ponec, J. Catal, 1995, 757,

431-438.

[90] M. Englisch, V S. Ranade, J. A. Lercher, Appl. Catal. A, 1997,163,

111-122.

[91] S. Galvagno, A. Donato, G. Neri, R. Pietropaolo, D. Pietropaolo, J.

Mol. Catal, 1989, 49, 223-232.

[92] C. G. Raab, J. A. Lercher, J. Mol. Catal., 1992, 75, 71-79.

[93] P. Claus, P. Kraak, R. Schödel, Stud. Surf. Sei. Catal, 1994, 108,

281-288.

[94] P. Claus, D. Hönicke, Chem. Ind. (Dekker), 1994, 62, 431-434.

[95] D. G Blackmond, A. Waghray, Chem. Ind. (Dekker), 1994, 62,

295-305.

[96] D. G Blackmond, A. Waghray, J. Phys. Chem., 1993, 97, 22,

6002-6006.

[97] E. Gindro, Dissertation Nr. 14714, ETH Zürich, 2002.

[98] H. Greenfield, D. S. Frederick, US Patent 3350450, 1967.

[99] E. Widmer, M. Casadei, Ro 86322, F. Hoffmann-La Roche, 1975.

Page 135

100] H. Ernst, Pure Appl. Chem., 2002, 74, 11, 2213-2226.

101] D. Rood, in "A Practical Guide to the Care, Maintainance and

Troubleshooting of Capillary Gas Chromatography Systems",

Wiley, New York, 1999, pp. 236-289.

102] C. J. Cowper, A. J. DeRose, in "The Analysis of Gases by

Chromatography", Pergamon Press, Oxford, 1983, pp. 225-261.

103] S. L. Morgan, S. N. Deming, J. Chromatogr, 1975, 772, 267-285.

104] "Organikum: Organisch-chemisches Gmndpraktikum", Johann

Ambrosius Barth Verlag, Leipzig, 1993, pp. 404-451.

105] J. F. Hayes, J. D. Hayler, T. C. Walsgrove, C. Wicks, J. Heterocyclic

Chem., 1996,33,209-212.

106] C. Aubert, C. Huard Penio, M. C. Lasne, J. Chem. Soc, Perkin

Trans., 1997, 7,2837-2842.

107] J. W. Wiench, L. Stefaniak, E. Grench, E. Bednarek, J. Chem. Soc,

Perkin Trans., 1999, 2, 885-889.

Page 136

Chapter 11

Appendix

11.1 Curriculum vitae

name

date of birth

citizen

Roc Novi

15.10.1977

Vignogn GR

Education

1984-1990

1990-1992

1992-1997

1997-2001

2001-2004

primary school in Savognin

secondary school in Savognin

gymnasium at the EMS in Schiers

diploma education in chemistry at the Swiss

Federal Institute of Technology in Zürich

Ph.D. thesis at the Swiss Federal Institute of

Technology in Zürich and the F. Hoffmann-La

Roche (VFCR) in Kaiseraugst

Page 137

11.2 Conference contributions

R. Novi, F. Rössler, P. Rys, "Hydrogénation of aliphatic nitriles;

thermodynamic and kinetic aspects", 6th International Symposium on

Catalysis Applied to Fine Chemicals, 6.-10. April 2003 in Delft, NL.

O. G Degischer, R. Novi, F. Rössler, P. Rys, "Application of a

hydrogénation catalyst modified by formaldehyde", 18th North

American Catalysis Society Meeting, 1.-6. June in Cancun, MEX.

Page 138

Date post:	13-Feb-2017
Category:	Documents
Upload:	truongtram
View:	224 times
Download:	0 times

Hydrogénation of Aliphatic Nitriles over Nickel Catalysts Modified by ...

Documents